CHAPTER3·METHODOLOGY

Is an intervention which relies on an online course to train

the client in cognitive behavioural techniques and on face­
to-face sessions to practise the techniques at least as
effective as face-to-face counselling in treating clients with
mild to moderate depression?
If a positive effect is realised from such an intervention, how
persistent might this effect be?
Is this intervention more likely to retain clients through to
completion compared to other interventions?
What are the potential savings of this intervention compared
to other interventions?

3.

2.

4.

(i) Introduction
The review of the literature in Chapter 2 provides direction to the analysis
of the study, which has given rise to the following research questions
namely:

1.

This chapter outlines the methodological considerations in the analysis
used to address these research questions and seeks to justify the chosen
approach. The next chapter provides the analysis of the results and
addresses the research questions.

The present study was undertaken within the writer's normal practice and
with referred, day-to-day clients. It was a feasibility study requiring further
studies to establish the effectiveness of the proposed combination of
online and face-to-face treatment of depression by the same practitioner.
Measurements of depression were taken before, during and after
treatment and also three months later. Clients participated well because
they understood and saw the relevance of the experiment to their own
needs. The results of the study are directly applicable to clinical practice
and represent a valid example of an integration of treatment and
research.

126



In the study, founded on a constructivistlinterpretivist paradigm,
quantitative and qualitative methodologies were used to reflect different
perspectives. Tools based on those derived from quantitative
methodology served to assess the efficaciousness of treatment from an
external and statistically demonstrable perspective. Case studies, derived
from qualitative interpretivist methodology, revealed a picture of changes
in the inner reality of participants as they progressed through their
treatment.

The quantitative component of the study lent itself more to experimental
than observational research because it relied on the application of an
intervention and measurement of before and after states.

The quantitative experiment was based on the comparative change
design, which is a quasi-experimental design. The design utilised before­
and after-intervention measures to assess change in outcome factors and
compared this change between treatment and non-random comparison
groups. The independent variable itself was not manipulated. Absence
of randomisation made this design vulnerable to alternative explanations
because of the possible effect of prior differences between groups. That
is, the difference between pre- and post-intervention might have been
partly or entirely due to confounding factors rather than the intervention
itself. Confounding factors included the underlying motivation to
undertake the study and the attraction of the novelty of online treatment.
It should be noted that random allocation by itself does not guarantee the
validity of estimates of causal variables. (Garson, 2003, Shadish, Cook &
Campbell, 2002, Heffner, 2005).

(ii) Participants
(a) Selection methodology criteria
Participants needed to meet the following criteria:
• suffer from symptoms of depression;
• have the capacity to access the Internet in private;

127



• have no currently active psychosis;
• not be receiving any treatment for their depression;
• be able to devote minimum 2 hours/week to the course;
• be willing to practise course advice and skills daily.

The criteria in sample selection aim to mirror qualitatively the population
which will most likely yield results directly related to the proposed
treatment. Unless participants suffer from depression there would be little
purpose in their participation given that the study is to test a treatment
approach. Unless they can access the Internet they would not be able to
undertake the online course. This does not imply that they must have a
computer exclusively dedicated to their use. In the study one participant
undertook the course in the writer's reception office. Any current
psychosis is highly likely to interfere with treatment and confound the
results. Receiving no other form of treatment assures as far as possible
that any improvement in mood has come about because of participation
in the intervention, that is confounding is minimised. The devotion of time
and willingness to practise are essential to any CST intervention because
of the need to learn and apply certain skills.

Individuals with mild to moderate depression were referred to the writer
for treatment by their general practitioners or other psychologists. The
writer saw each referred client individually and offered the opportunity to
participate in an experimental treatment procedure which consisted of a
mix of an online course on CST skills and face-to-face treatment.
Interested persons were advised that the treatment would be of no cost to
the participants. The participants would undertake the course on their
own and contact the writer by email approximately weekly. If they felt the
necessity for face-to-face sessions to address particular difficulties in the
course or any other issues interfering with their capacity to pursue their
treatment they could freely arrange attendance with the writer.

The election of participants to have a face-to-face session may have a
confounding effect on the treatment effectiveness of online methods.

128



However the effect is hypothesised to be small because the sessions
were meant to help individuals overcome obstacles to their online
learning rather than replace the online modules with traditional treatment.

(b) Participant characteristics
The level of depression was assessed by the self-reporting BOI and by
the clinical interview-based HORS. Participants were selected on the
basis that BOI scores showed depression at the mild to moderate levels
namely below the BOI score of 29 and above 13 (Beck, 1996). In total
nine participants were chosen.

All participants spoke English fluently. One participant was a Middle-East
non-resident awaiting protection visa approval. He was unemployed and
supported by his brother. All were at professional/managerial
occupational levels. Six participants completed all phases of treatment.

The three participants who left the course completed some modules and
attended one or more sessions. Of those who withdrew one stated that
her depression had lifted when she married, another that she improved
after completing part of the course and the third that he was undertaking
the course on his own and applying skills in his day-to-day functioning.

Gender and educational levels for completers are outlined in Table V and
for non completers in Table VI:

Table V. Completers - ethnicity, gender, educational level, age
Participant Ethnicity Gender Highest education Age
Aaron Anglo-Celt M Year 12 47
Bill Anglo-Celt M Tertiary (15) 28
Cathy Asian F Tertiary (15) 26
Dee Anglo-Celt F Year 10 31
Eric Mid East M Year 12 26

Fred Mid East M Tertiary (15) 35
Average Mode Anglo-Celt Mode Male 13.16 32.17

129



Table VI. Non completers - ethnicity, gender, educational levels, age
Participant Ethnicity Gender Highest education Age
Glenda Anglo-Celt F Tertiary (15) 28
Harry AnQlo-Celt M Year 12 55
lisa Anglo-Celt F Year 10 35
Average Mode Anglo-Celt Mode 12.08 (years) 32.67

Female (years)

An analysis of the information in Tables V and VI revealed that all non­
completers were of Anglo-Celtic background (two female and one male)
of whom one was tertiary-educated (15 years schooling). Completers
represented both genders, ranged in age from 26 to 47 and were from
three different ethnic backgrounds. Their educational level varied from
lower secondary to tertiary. All participants had depression at the upper
half of the moderate level as measured by the BOI, except for Glenda
whose depression was at the lowest end of the moderate level.

Table VII shows no statistically significant differences between age,
gender, ethnicity and education between completers and non completers.

Table VII. Significance of ethnicity, gender, educational levels, age
between completers and non completers
Group Ethnicity Gender Average years Average

education age
Completers 6 Anglo-Celt 3 Male 4 13.16 32.17
Non completers 3 Anglo-Celt 3 Male 1 12.08 32.67
chi squares chisq chisq t test t test

=0.56 ns =0.056 ns =0.52 ns =0.98 ns

(iii) Procedures
(a) Study design
The study design is outlined below and illustrated in Figure 1. An
individual session was held for each participant, which began with the
testing of the level of depression with the BOI and the HORS. After the
level of depression was confirmed as within the bounds of the study (mild
to moderate) an explanation of the treatment program was provided. The
participant was informed that he or she would first undertake the online

130



course, beginning with the first module and then continuing to advance
module by module. There were 14 modules. The modules presented the
knowledge to be imparted and provided further material for enrichment
purposes through links to other sites. Relaxation sites including games,
humour and music were also linked in order to reward the participant's
efforts with enjoyable activities. Participants were encouraged to visit the
sites and enjoy the offerings. At the end of six modules there were
exercises requiring the application of the knowledge imparted in the
module to practical day-to-day situations in the life of the participant. For
example the exercise at the end of the third module required the
participant to write examples on the application of the concept of looking
at things differently. Seven modules required the completion of the
Goldberg depression test which was situated in a linked site, in order to
give the participant feedback on the current level of depression and
hence level of progress being made. Nine modules provided the
participant with a simple mood monitor for additional feedback.

Participants were informed that the course was individualised to the
extent that there were no timetable requiring adherence and that the
website could be accessed at any time. They would move from module to
module at their own pace. They were invited to communicate any
difficulties or other issues they wished by email and to arrange for face­
to-face sessions to address 'stuck points' or where their depression
deteriorated to the extent that their ability to apply the skills learnt became
impaired. No limit was imposed on the number of sessions that a
participant could have because of the individualised nature of the course
itself. Participants were advised that following completion of the online
training course they would attend a session to complete the SOl and the
HORS. If the SOl or the HORS indicated that depression continued at
clinical levels participants would be invited to undertake face-to-face
sessions. After the sessions and also three months after completion of
the sessions they would be tested with the SOl and the HORS. If
depression continued the participant and the writer would discuss
alternative actions.

131



(b) Ethical practice
The intending participants were assured that they were unlikely to
experience risks, inconvenience or discomfort. At worst, their depression
might not get any better or it may worsen, in which case they were to
consult the writer who would then refer them to their respective GPs to
arrange for alternative treatment outside the study. If treatment itself
raised distress or other issues and participants felt the need for
counselling outside the project, they could contact the telephone
counselling service or visit the local community health centre. Intending
participants were advised of the likely length of the project of between 4
and 6 weeks, that the project was entirely voluntary and they were free to
leave at any time. One-on-one treatment would take another 6 to 9
weekly hourly sessions. A recall would take place three months later.
The total length of time of the intervention was expected to be about six
months. Individual participation was estimated at between 11 and 16
weeks.

Intending participants were assured that they could not be identified
through their test results or other information provided. The data would
be held in a locked cabinet and in password-protected computer storage.
No access would be allowed to anyone other than the writer or his thesis
supervisors and five years after the project all data would be destroyed.
Intending participants were advised that the project had been approved
by the Human Research Ethics Committee of the University of New
England (Approval No HE05/192 valid to 26/09/2006)
Ethical issues are embodied in the document required to be signed by
intending participants. No concern was expressed by any participant at
any stage. A copy is presented in Appendix (i).

Figure 1. Study design

132



lnitialintervtew
Completes
BOI, HORS

Begins ontine
course

Progresses with
course; emails
Goldberg tests
Mood monitors

Arranges and
attends

session(s)

Attends session
Compl&t&s
BOI,HORS.

Questionnaire

Arranges and attends
sesmon(s): completes

questionnaire

")(===:~OiscusS course of
;" actioo

Analysis of case
studies

133



(iv) Instruments
Quantitative research in the study utilises psychological tests, rating
scales and Likert questionnaires to gather data. A caution applies. The
sample in the study is too small to allow for statistically sound
comparisons to be made with other studies. Comparisons need to be
interpreted with caution and any implications which follow should be seen
in this light. That is, the results of the study point to the merit of further
research using control groups and larger samples, rather than establish
the efficacy of the treatment approach used in the study.

(a) The BOI and the HORS
The SDI and the HDRS were applied at four stages, namely at the start
and the end of the course, at the end of the sessions and at follow up.
The writer chose a cut-off of 9 in the SDI because in his experience
clients with lower scores rarely sought treatment, stated that they felt
depressed or showed symptoms of clinical depression. Seven
administrations of the Goldberg Depression Scale and nine completions
of a mood monitor averaged over the previous three days at morning
noon and evening (10=best mood, O=lowest) throughout the course
provided participants and the writer with feedback on their progress.
Participants' views of the course and its usefulness were assessed by
Likert Scales on three occasions: on completion of the course, following
the sessions and three months later. The questionnaires are presented
in Appendix (ii).

The research questions were tested against similar results in five studies
of face-to-face CST treatment of depression using the SDI as a measure.
One was a metastudy of twelve depression studies of which nine had
been chosen on the basis that they used CST and provided data which
permitted comparison with the present study. All studies were selected
because they included a wide variety of CST approaches enabling the
present study to be compared with a range of different treatments and
methods. For example some studies used manualised treatment, namely

134



the procedures to be followed were described in a way which facilitated
their application in particular cases not unlike a technical manual.

Data from the depression measures were analysed by means of tests of
the difference between means and distributions of scores of the
participant group at different stages of the intervention and between the
participant group and similar groups in other studies.

The caution stated earlier needs to be reiterated and expanded. Apart
from the small sample in the study there are many other variables
involved. For example the comparison studies did not indicate the level of
experience of the therapists nor how they applied CST during the
treatment process. There was also little information on whether or not
participants had prior experience in being treated with CST. All these
issues imply that the present study rather than establish the efficacy of
online treatment with face-to-face support, points to the need for further
research using control groups and larger size samples.

(b) Likert scale questionnaires
The Likert method was adopted for the participant questionnaires
because it is universally used, questions are easy to answer and analysis
is speedy. The questionnaires were analysed to identify participant views
on the online course - presentation, the usefulness of skills taught and
the ability to transfer skills to day-to-day functioning.

Quantitative data were obtained by questionnaires composed of 5-point
Likert Scales. Completers responded to questionnaires (Appendix (ii))
during the evaluation of each phase, namely the course, the post-course
sessions and the follow up. At the end of the course two different
questionnaires were used. The first collected completers' views on the
course itself - quality of communication, enrichment material, exercises,
evaluation, level of challenge and perceived availability of their
psychologist. The second questionnaire concerned the extent to which
tasks taught through the course were transferred to the individual's day-

135



to-day functioning. This questionnaire was repeated at the end of the
post-course sessions and the follow-up. The questionnaires were
completed face-to-face or over the phone in the case where the client
could not attend. Clients were invited to expand on their responses and
provide any other information they wished on the intervention.

In the present study emails and different completed tests and
questionnaires were considered documents. Tests informed on the
extent of the disorder, progress in getting better and post-treatment
resilience. Emails indicated progress with the course and also provided
the participant with a ready and convenient means to communicate on
the course with the writer. Questionnaires gave participants the
opportunity to inform the study on the short-term and long-term
effectiveness of the application of skills learnt in the course in addressing
depression. The interviews at the three different stages (pre-course,
post-course and follow up) were open ended and participants were given
full scope to comment on the effects of their disorders on day-to-day
functioning, the usefulness of areas taught in the course and their overall
views and ideas on the course. Triangulation was demonstrated where
test scores, questionnaires and interviews consistently provided the same
message.

The writer, as the treatment provider, is a participant in the events being
studied. He is responsible for the selection of treatments, the assessor of
their effectiveness as well as the evaluator of the impact of his own work
on his own clients with no independent data analysis. This does affect the
validity of the study and it needs to be borne in mind that the study tests
the feasibility of a combined face-to-face and online course intervention in
the treatment of depression. Any wider application of the approach
should be established by larger scale studies appropriately designed with
control conditions.

136



(c) Case Study data sources
The case study method was used to trace the lifting of depression of
individual participants as they progressed with the intervention. A single
case study design was chosen because it enabled the writer to share
however imperfectly the participant's private world as he or she struggled
through depression and on to a state of wellbeing. The richness of
material provided in the case studies cannot be gained through reflecting
on discrete test scores.

The single case studies were used to trace the lifting of depression of
individual participants as they progressed with the intervention. These
were informed by individual interviews, emails, calls and Likert
questionnaires. The initial interview provided information on:
• life history - family, work, future plans;
• history of depression and events that serve to maintain it
• history of any treatment to date;
• attraction to undertake the intervention particularly the online

component;
• Internet capabilities - verification that the participant has online

access.

The midcourse emails and calls provided information on progress
measures, course exercises and any issues relevant to the course. Mid­
course sessions informed on 'stuck points' which impeded progress,
setbacks to improvements and flagging motivation to pursue with
treatment.

The end-course interviews informed on:
• Participant perception of the extent to which depression had been

defeated since beginning the course;
• any remaining clinical issues which required addressing;
• the component of the course which proved to be the most helpful;
• changes in behaviour, that resulted from the course, for example

the incorporation of thought challenging in daily behaviour.

137



The post-course sessions addressed clinical issues and generally aimed
to free the individual from depression.

The post-three month interviews informed on:
• robustness of the treatment, that is the ability of the individual to

maintain a depression-free state in the intervening period;
• any relapses and how they were addressed;
• behaviour changes and their incorporation into the individual's d

aily functioning.

(v) Program
(a) The research environment
The research environment consists of the methods of participant - writer
communication and the online course.

The email was the common mode of communication for participants on
the online course. Participants notified their mood monitor, test scores
and exercise results by email to the writer. Emails were used as a
general form of communication between the individual participant and the
writer. Participants notified difficulties encountered in the course, mood
relapses, questions and observations by email. The writer ensured that
he responded as quickly as possible and within a few hours. There are
many reasons for aiming to ensure a rapid response. They include:

• maximisation of reinforcement to the participant by reducing the
time interval between question and response;

• resolution of impediments to progress;
• validation of the individual's concerns;
• the strengthening of a caring environment.

138



In a face-to-face environment these aspects would be provided virtually
instantaneously. In online treatment of the kind in the study they need to
be established through emails.

Other modes of communication included face-to-face at the briefing stage
of the course during which participants were given user name and
password access to the website with the online course
(www.consultantpsychologist.com/treatment. username: client, password:
module). Further face-to-face communication took place in any treatment
sessions provided and in the evaluation sessions at the end of the online
course, the sessions and at the follow-up.

(b) The online course
The online course is based on CST. The mode of presentation of the CST
concepts is hypothesised to differ between practitioners. The mode also
differs between a practitioner and anyone particular client because no
two clients can be expected to have identical issues underlying their
respective depressions or learning capacities. The presentation may
change over time as the treatment providers reflect over their approach,
consider client suggestions and incorporate new material learnt through
professional development and other learning activities.

The course consists of a series of modules and is constructed on the
basis of the writer's CST notes which he uses for face-to-face treatment
of depressed clients as modified over years of practice. The completion
of one course module leads to the beginning of another. Modules consist
of webpages to be read. Certain modules end with exercises for
submission to the writer. The webpages refer the participant to other
sites with enrichment material on the contents of the module, games and
humour. The information forwarded by the participants includes mood
monitor scores for morning, midday and evening for each day averaged
for the last three days and scores attained by them on the Goldberg
Depression Scale. There may be a practice effect due to the requirement
to complete the Goldberg a number of times, however the participant is

139



motivated to complete the test realistically given that its main purpose is
to provide feedback. In any case the mood monitor provides an
independent measure of assessing the current level of depression should
the Goldberg score be overly optimistic. Participants access the test on
an Internet site (http://mood-monitor.perbang.dk/index.php?aid=35) which
is referred to at the end of seven of the modules. The scores give the
participant and the writer feedback on any mood changes and the extent
of depression symptoms which exist at that time; that is both measures
provide independently an estimate of treatment progress and hence
reinforcement.

The course modules are:

• an introduction and explanation of procedures - the first Goldberg
test result to be emailed to the writer;

• looking at things differently - a discussion on Epictetus;
• practising looking at things differently, exercises on the application

of the Epictetus principle (people react to events the way they
interpret them), mood monitor scores and Goldberg test results to
be completed for submission to the writer;

• formation and management of stress - knowledge test, mood
monitor score to be submitted to the writer;

• stress symptoms and their relationship to the survival system;

• continuation of presentation of stress symptoms; their relationship
to the survival system; completion of exercise based on personal
experience of a recent stressful situation, mood monitor score and
Goldberg test result for submission;

• relaxation (I) - breathing relaxation;
• relaxation (11)- PMR (progressive muscle relaxation), other forms of

relaxation, diet;

• basic meditation technique, test on experiences in the application
of meditation, mood monitor score and Goldberg test result for
submission;

140



• introduction to CST-Socratic questioning, Aristotelian logic,
behavioural exposure, test on the CST model based on recall of
recent experience of depressed mood and associated thoughts
and behaviour, mood monitor score for submission;

• CST key concepts - core beliefs, conditional rules, automatic
thoughts, linking thoughts mood and behaviour, exercise on the
experience of application of CST through exposure by returning to
a place avoided since becoming depressed, mood monitor test
score and Goldberg test result for submission;

• CST tools (I): finding automatic thoughts, identifying irrational
beliefs, discussing two of the rules of Aristotelian logic; exercise
similar to the one earlier on application of CST principles through
exposure to a second behavioural experience and mood monitor
score for submission;

• CST tools (II): discussing remaining rules of logic;
• analysing thoughts - common negative thoughts of depressed

individuals; exercise on analysing negative depressive thoughts,
mood monitor score and Goldberg test result for submission;

• strategy summary.

(c) An illustration of a course module
The third module "Practising looking at things differently" is discussed to
illustrate the course. The module itself is in Appendix (iii) and should be
referred to for a better understanding of the following discussion. (The full
course can be accessed on www.consultantpsychologist.com/treatment
username "client" password "module").

The module provides examples and practice on the Epictetus principle,
namely "people react to events the way they interpret them and not how
they are". The module begins with four everyday examples of the
principle. The examples are to emphasise the practicability of the concept
and its application to day-to-day functioning of the individual. The client is
led to see the link between thought and mood by the examples and is

141



given ample opportunity to practise the skill of looking at things differently
through the use of illusions and by an assignment for submission to the
writer. The assignment requires the presentation of four different
examples of looking at things differently which reflect changes in mood.
The illusions or ambiguous drawings are on two different sites which are
hyperlinked to the module. The writer has found through experience that
clients react positively to ambiguous drawings and can transfer the
application of the skill of looking at things differently to their own day-to­
day life. The assignment itself requires the completion of the mood
monitor and the submission of Goldberg test results.

Depression is characterised by low concentration and fatigue. These
symptoms are addressed by inviting the client immediately after
completing the assignment and the mood monitor to:

• Reflect on a waterfall scene. The client is asked to imagine that it
is a warm summer's day and that he or she has some spare time.
The client takes a trip to the bush, comes to the waterfall and tries
to imagine how relaxing it is;

• Listen to two well-known Scott Joplin tunes, Bethena and The
Entertainer whilst reflecting on the waterfall scene. Both tunes
should have a wide appeal amongst the different age groups. The
tunes are hyperlinked to sites with additional tunes which might
further attract the client. They provide a pleasant interlude which is
itself of therapeutic value;

• Play at a game site. The site was chosen because the games are
simple to master and playing them provides pleasure;

On completion of the module the client is encouraged to go for a walk to
the park or the shops, talk to a friend on the phone, work on a hobby or
get on to some of the tasks that are waiting attention. In this way the
cognitive and the somatic components of depression are addressed. The
client is then ready to begin the next module.

142



CHAPTER 4 - FINDINGS

The previous chapter discussed the methodology of the study which
involved a mixed methods approach incorporating quantitative surveys
and qualitative case studies. This chapter aims to provide answers to the
research questions posed in this study, namely:

1. Is an intervention which relies on an online course to train
the client in cognitive behavioural techniques and on face­
to-face sessions to practise the techniques at least as
effective as face-to-face counselling in treating clients with
mild to moderate depression?

2. If a positive effect is realised from such an intervention, how
persistent might this effect be?

3. Is this intervention more likely to retain clients through to
completion compared to other interventions?

4. What are the potential savings of this intervention compared
to other interventions?

There are three sections in the chapter. The first section draws on data
from quantitative and qualitative study components to address the
questions. The second section presents the case studies which prOVide a
descriptive background to the quantitative results, that is, they reveal
changes to an individual's functioning as depression lifts. The third
section evaluates the individual components of the intervention.

Normality of the BOI, HORS, mood monitor averages and Goldberg
scores was first tested. The W/S test was used. This test estimates the
significance of the difference between a frequency distribution based on a
particular sample and a normal or Gaussian distribution (Kanji, 2001,
p.65). The results reveal that the ratio of standard deviation to range
varies from 2.3 to 3.0 and in all cases lies within the upper and the lower
boundaries of the 0.025 level of significance. The null hypothesis, that
the underlying distribution is normal, is accepted. This result enables the
valid use of parametric statistics.

143



(i) The findings of the present study
Nine participants, six men and three women, were chosen in the present
study. The participants underwent an initial interview during which their
depression was assessed by the BOI and the HORS. They were informed
that the treatment approach consisted of a mix of an Internet course and
face-to-face sessions. The course included periodic assignments and
depression test assessments which had to be reported to the writer by
email. Face-to-face sessions could be arranged by the participants with
the writer on the basis of their needing remedial help or if it was clear that
the course was not benefiting them. Following the end of the course face­
to-face sessions could also be arranged by participants to practise the
application of skills learnt during the online component of the intervention.
Assessment sessions were included at the end of the course, the end of
face-to-face treatment where this continued after the course, and three
months after the end of treatment. The assessments included the BOI
and the HORS as well as Likert scale questionnaires measuring
participant satisfaction with the course and treatment as a whole.

Six participants, four men and two women ranging in age from 26 to 47
with a mean of 32.2 years and a standard deviation of 8 years, completed
the course. The other three completed part of the treatment and managed
to overcome their depression with the aid of principles learnt through the
course. Participants took from 26 days to 153 days to complete the
course, with an average of 86.5 and a standard deviation of 44.8. They
needed from 2 to 10 sessions each, averaging 4.7 with a standard
deviation of 3.2. Some of the endpoint assessments were undertaken by
phone.

The following table summarises participant performance. The difference
between pre and post treatment in the BOI is significant, indicating the
success of the intervention in assisting participants to reduce levels of
their depression. Further details are provided in subsequent tables.

144



Table VIII. Summary results of current study
Statistic Pre-treatment Post-treatment Follow up Significance

(course and
sessions)

No. participants 9 6 6
in study
BDI Mean, (Std 25.8 (10.27) 5.0 (3.2) 4.8 (1.5) t-test pre/post
Deviation) treatment: p<.01
% with BDI<9 * 0 83 100
* considered in the literature as signifying the cut-off for subclinical
depression. See the note on the SOl cut-off score in Chapter 3 (xvii).

(ii) Comparison studies
The results from the present study are compared with similar
results from face-to-face and online studies in order to
demonstrate relative effectiveness. The comparison studies were
chosen on the following basis:

• the intervention aimed to treat depression either solely or with
other psychopathologies;

• the treatment method was based on CST;
• the study was reported in English language mainstream journals;
• control groups were used;
• the main measure for the face-to-face comparison studies was the

SOl. (at the time of writing insufficient online interventions were
found which used the HDRS);

• the study included follOW-Up;
• for the online studies the material was presented in modules.

The salient features of the comparison studies are summarised in the
table following.

145



Table IX. Key features of chosen studies
Study authors
(i) face-to-face studies
Cahill et al. (2003)

Westen & Morrison
(2001 )

Shapiro et al. (1994)

Jacobson et al.(1996)

Jacobson et al.(1991)

(ij) online studies
Andersson et al. (2004)

Christensen et al. (2004)

Clark et al. (2002)

Marks et al. (2003)

Description

SOl scores were collected at each session (1) to describe
the outcomes of clients who received CST for depression,
and (2) to compare client outcomes for completers and
non completers. The study supported the effectiveness of
CST.
A meta-analysis component of 12 studies on the efficacy
of manualized evidence-based (CST) psychotherapies for
depression found that treatments yielded significant short­
term effects (on the SOl) which continued at follow-up.
Depressed clients completed manualized treatment (CST
or IPT). In the study the SOl showed an advantage for
CST though both methods were effective, irrespective of
the severity of depression or the duration of treatment and
follow up.
The study set out to test effectiveness of behavioral
activation and teaching of skills to modify automatic
thoughts (AT) as components of CST. The study found
that these components were as effective in reducing
depression as the full CST treatment both at the
termination and the follow up stages.
The study compared CST and marital therapy (SMT), and
a treatment combining SMT and CT in the alleviation of
wives' depression and the enhancement of marital
satisfaction. CST was more effective than SMT for
depression in maritally non-distressed couples, whereas
for maritally distressed couples the two treatments were
equally effective.

Participants were assessed for depression on the website
and independently randomised to the treatment or control
conditions. The material was presented in modules which
ended with a questionnaire. Responses were
automatically sent to a therapist who gave feedback by
email. There was a significant reduction in mean scores
of the treatment group but not the control group.
Statistically significant improvements were found between
pre-treatment and follow-up.
This was a randomised control study evaluating the
efficacy of two online interventions (SluePages,
MoodGYM) for depressed individuals. Participants were
recruited from survey responders assessed as clinically
depressed. They were allocated randomly to SluePages,
MoodGYM and a control group. The experimental groups
showed improvements over the control group.
The intervention consisted of an online interactive
depression self-paced and self-help skills training
program on the use of cognitive restructuring techniques,
without input by treatment professionals. The Internet
program was evaluated by comparison with usual care in
a health maintenance organisation. The experimental
group was gender- and age-matched with a similar sized
non-depressed group.
The study evaluated a program of computer-aided self­
help for clients in a primary care clinic. Patients
undertook an online CST program (Cope) with interactive

146



voice response assistance. There were significant
improvements for experimental group participants.

Wright et al. (2005) This was a waitlist control group study which compared
the efficacy of computer-assisted CST with face-to-face
CST. The study found no significant differences on these
two measures between either modes of treatment both of
which were superior to the control group

In summary the key features illustrate the effectiveness of CST in face-to­
face or self-study formats in reducing depression. An examination of the
research questions follows.

(iii) Research question 1:
Is an intervention which relies on an online course to train the client in
cognitive behavioural techniques and on face-to-face sessions to practise
the techniques at least as effective as face-to-face counselling in treating
clients with mild to moderate depression?
Relative effectiveness was assessed by comparing the present study and
the comparison studies in relation to (i) rates of treatment success, (ii)
times taken to complete the intervention and (iii) numbers of sessions
provided through the intervention.

The following table presents data for several measures for pre-treatment,
post-treatment and follow-up for the present study and the comparison
studies and enables comparisons to be made. All of the comparison
studies have been discussed in Chapter 2.

In the table the two Jacobson studies and the Cahill study use essentially
CST as the methodology in face-to-face treatment. The Westen and
Morrison (2001) and the Shapiro et al. (1994) studies utilise manualised
treatments in a face-to-face presentation mode. In manualised
treatments the provider follows prescribed directions in delivering therapy,
a similar process as teachers' lesson guides. Manualised treatment aims
to standardise presentations in order to improve quality particularly for
less experienced providers. The present study is based on an Internet­
delivered course aided by face-to-face sessions where required.

147



Table X. Pre-treatment, post-treatment and follow-up
Study Cahill et Westen & Shapiro Jacobson Jacobso Present

al. Morrison et al. et n et study
(2003) (2001 ) (1994) al.(1996) al.(1991 )

No. began 58 336 (9 CST 150 56 8 9
treatment studies)
Pre-treat SOl 28.38, 24.5 29.8 29.3 25.8
mean (SO) = (10.27) (6.34) (6.3) (7.6) (2.5)
Average no of 11.6 8 20 20 3.83
sessions per
participant
No completing 35 237 (est) 117 50 7 6
% completing 60 71 78 89 87 67
Post-treat SOl 10.30 10.14 9.48 10.1 7.4 5.0
mean (SO) = (10.22) (7.15) (7.72) (9.6) (6.7) (3.2)
Post-treat % 48.9 55 est 44.0 45.0 est 29.6 est 83
with SOl < 9
Post-treat 1.76 est 2.2 est 1.77 est 2.48 est 3.27 est 7.57
Effect size
Follow-up no 3 6 3
of months
Fallaw-up 115 47 6
n=
Follow-up 7.8 (8.2) (all 10.15 10.3 4.8
SOl mean depression (8.69) (8.6) (1.5)
=(SO) studies)
Follow-up 50.1 est 46 42 est 100
% with SOl <
9
Follow-up 18 months 3 months 6 months 3 months
term - 2 years
Fallaw-up 0.49 2.65 est 438
Effect size
t-test means t=17.04 t=12.33 t=5.93 t=13.54
pre/ post-treat p<.01 p<.01 p<.01 p<.01
t-test means t =0.60 t =0.11 t = 0.12
post! follow-up p>.05 ns p>.05 ns p>.05 ns

Notes
1. Entries with "est" are estimated from other data provided in the

study.
2. Numbers with SOl <9 are estimated by assessing areas under the

normal curve using the means and standard deviations provided.
3. The Shapiro et al. (1994) study does not provide separate SOl

data for CST and interpersonal therapy. Combined data are
presented in the table because that study finds no significant
differences between the two methods.

148



4. Blank cells indicate that no comparable data was given in the
study.

The table presents pre-, post- and fOllow-up measures for the present
study and the comparison a range of comparable studies which enable
comparison of outcomes. The comparison is made difficult because a
sufficient number of studies which use the same measures and
parameters were not found. Reliance had to be made on studies using
some of the same measures to those in the present study. The table
indicates that the mean SOl score of participants at pre-test varies within
a band of less than 6 points clustered around the upper cut-off of the
moderate category of 20-28 (Beck, 1991). On completion of treatment
the mean moves to below the mild lower cut-off of 14. T-tests between
pre-treatment and post-treatment means for the three comparison studies
and the present study find the differences significant indicating the
effectiveness of the treatments (17.04, 12.33, 5.93, 13.54, p<.01).

The t-tests of the difference in proportions of completers with BOI<9 at
post-treatment in the present study and in the comparison studies
combined in a group (by multiplying for each comparative study the
proportion of those with BOI<9 by the number of participants, totalling
these numbers and finding the percentage of the totals with BOI<9
against the overall total number of participants) reveal that the difference
is statistically significant (t=6.89, p<.01). This leads to the conclusion that
the present study has a significantly higher proportion of completers with
subclinical depression at post-treatment than for the combined
comparison studies.

(a) Time taken to complete treatment
Assessing effectiveness requires taking into consideration the length of
time to complete the intervention in order to minimise any contamination
effects that may arise. For example an individual might have had an
unduly positive or negative life event occur, which independently affects
mood such as a marriage or death in the family. In the present study one

149



of the non completers (lisa) married just after she began on the course.
On return from her honeymoon she did not continue with the course
because she felt that marriage had markedly improved her life and
resolved her depression. The following compares times taken to
complete the interventions in this study and in the comparison studies.

Table XI. Average number of days per participant to complete face-to­
face interventions

Shapiro et al. (1994) 56
Jacobson et al. (1996) 140
Jacobson et al. (1991) 140
Mean (SD) 112 (48)
Present study 88

The mean number of days for completion of the comparison studies is
112 with a standard deviation of 48. The mean of 80 days for the present
study is similar. It lies within the range of the mean plus one standard
deviation.

The time taken to complete the intervention in comparison with other
online treatment studies is one indicator of its relative standing amongst
peers. Data for the present study reveals that the course takes no longer
to complete than most of the other Internet interventions. In the present
study the mean length of time for completion of the course and sessions
is 88 days. The following table indicates the days taken on average to
complete treatment in the Internet studies including the present study.

Table XII. Mean number of days to complete Internet interventions
Andersson et al. (2005) 70
Christensen et al. (2004) 42
Clark et al. (2002b) 224
Marks et al. (2003) 65
Mean, SD of studies 100,83
Present study (course plus sessions) 88

150



The table indicates that the mean number of days for completion of the
four comparison Internet studies is 100 with a standard deviation of 83.
The present study took 88 days to complete. This difference is not
statistically significant. It lies within the range of the mean and one
standard deviation.

(b) Support
Support was made available to participants through sessions, email and
on very few occasions by phone, principally to enquire as to lack of
progress. The number of sessions delivered is a key factor in assessing
effectiveness. An intervention that achieves subclinical SOl scores in 20
sessions is defined to be less effective than one achieving similar results
in fewer sessions.

The following table outlines the support provided to participants. The
support consisted of responses to emailed exercises, and emails and
calls generated or responded to by participants and the writer as the
treatment professional on topics such as comments amplifying mood at
the time. The time estimated for each task is included in order to highlight
the support provided in relative terms. The table indicates that close to an
hour was devoted on the average for every participant.

Table XIII Support: number of tasks and estimated time devoted to emails
and calls to participants
Support Completers Non completers Total

Tasks Time (est) Tasks Time Total Total
(est) tasks time

(% total
time)
(est)

Emailed exercises and 54 6.30 hrs (7 6 0.50 hrs 60 7.00
mood monitor scores (9 per mins each) hrs
participant) and Goldberg (74.3%)
scores attended to by the
writer (7 per participant)
Emails by participant or 8 0.80 hrs (6 7 0.70 hrs 15 1.5 hrs
responded to by the writer mins each) (15.9%)
Calls by participant or 6 0.50 hrs (5 5 0.33 hrs 11 0.91
responded to by the writer mins each) hrs

(9.7%)
Total 68 7.60 hrs 18 1.82 hrs 86 ~.42 hrs

151



The table illustrates that almost three quarters of support time was
estimated to have been devoted to email processing in relation to the
online course requirements. Other email support took up 160/0 and phone
support took up less than 100/0.

The implications of support appear to be:
• Requirements by the treatment professional to respond to emails

or generate emails, and for any calls that may have to be made or
responded to account for about an hour for each participant. Three
quarters of this time is devoted to assessing and responding to
exercises in the online course. Support devoted to calls totals less
than one hour for all participants and is hence small;

• the number of sessions indicated in the table include one hour per
participant for support. If this is considered the number of face-to­
face sessions reduces to below 3.

The number of sessions delivered is a key factor in assessing
effectiveness. An intervention that achieves subclinical SOl scores in 20
sessions is defined to be less effective than one achieving similar results
in fewer sessions.

A schedule of weekly sessions represents the usual practice and is
exemplified in studies by Cottone et al. (2001) and Gaynor (2003).
Whitfield (2003) observed that traditionally CST sessions were provided
weekly, though there was little evidence to support this frequency, and
hypothesised that access to treatment may be improved by decreasing
the number of sessions through the use of self-help.

Table XIV presents the number of sessions delivered in the comparison
and the present studies. The hours in the present study were increased
by one for each participant (total 9) to account for the time devoted to the

152



tasks such as email support as explained earlier. In the Jacobson studies
the face-to-face sessions were a week apart.

Table XIV. Sessions provided to all participants and completers in face­
to-face studies combined and the present study

Comparison studies Present study
No of participants 272 9
No of sessions 3152.8 34.5
Mean sessions per 11.59 3.83
participant
No of completers 209 6
Mean sessions per 15.09 5.75
completer

Fisher exact tests indicate the differences in the ratios of sessions to
completers and in the ratio of sessions to participants for the combined
studies and the present study are significant (p<.01 for both tests). The
table above demonstrates that the average number of sessions in the
present study per participant is 620/0 less than for the face-to-face studies.

(c) Conclusion
The data presented in the tables in this research question indicate that in
relation to the comparison studies:

i. the participants in the present study began with similar levels of
depression;

ii. the present study at post-treatment achieves a significantly lower
mean BOI score;

iii. the present study at post-treatment achieves a significantly higher
proportion of completers with BOI<9;

iv. the present study does not differ statistically significantly in the
mean number of days for completion;

v. the differences in the ratios of sessions to completers and in the
ratio of sessions to participants for the combined studies and the
present study are significant;

vi. the time devoted to email support appears to be one hour for each
participant.

153



In the light of these findings the first research question would seem to find
some support from the results of this study. The intervention in the study
appears to be as effective as face-to-face counselling in treating clients
with mild to moderate depression in CST skills.

(ii) Research question 2
If a positive effect is realised from such an intervention, how persistent
might this effect be?
It is first necessary to gain an appreciation of progress through the
intervention. This is provided through different vantage points and is
amplified by referring to the case studies.

(a) Progress measured by the BDI and the HDRS
One view is provided in the following table, which shows means and
standard deviations for the SOl and the HDRS at the pre-treatment, post­
course, post-sessions and follow-up stages.

Table XV. SOl and HDRS means and standard deviations at each
intervention phase

Phase SOl SO HORS SO
Mean Mean

Pre-treatment - intend to treat (n=9) 25.78 2.49 26.00 5.39
completers (n=6) 26.50 1.50 25.50 3.00

Post-course (completers) 4.7 3.1 4.3 1.6
Post-sessions (completers) 5.0 3.2 4.0 0.8
Follow-up (completers) 4.8 1.5 4.2 1.5

The table reveals that the pre-treatment means for completers and non­
completers lie in the upper half of the BDI moderately depressed category
(20-28) (Beck et al. 1996) and similarly for the HDRS (14-30) (Hamilton
1960, Bagby et al. 2004). The measures indicate that at pre-treatment
compieters were relatively more homogeneous in relation to their SOl
scores compared with all participants. The standard deviation widens for
the BDI in the post course and post sessions phases of the intervention
as the scores move away from congregating at the high end of the

154



moderate category and increasingly spread within a wider range of the
non-depressed category. At follow up the range becomes narrower as
the mean is at its lowest value.

Another view of progress through the intervention seen in terms of the
SOl and the HORS is provided in the following table. This table lists t­
tests for SOl and HORS means of completer data at different phases and
indicates the significance in changes in mood as participants advance
from pre-treatment to follow-up.

Table XVI. t-tests for SOl and HDRS means of completer data at
different phases

Comparisons SOl HORS
1 Pre-treatment - post-Internet course t=14.85 p<.01 t=13.15 p<.01

2 Post-course - post-individual sessions t=-0.54 ns t=-1.00 ns
3 Post-individual sessions - follow-up t=0.12 ns t=1.35 ns

4 Pre-treatment - follow-up t=65.0 p<.01 t=17.35 p<.01

The results indicate a significant reduction in means on both SOl and
HDRS before treatment and after the online course, and before treatment
and three months after completion. The implication is that the
intervention can be considered to be successful in assisting participants
to reduce their depression significantly following completion of the course
and to maintain the gains that have been made three months after
completion of the sessions. No significant differences emerge in either
test administered between the end of the course and the follow-up. The
writer hypothesises that the major reason is related to the fact that four of
the six completers required no sessions after completing the course. The
Internet course worked fully for them. That is their treatment needs were
met fUlly through their efforts on the course. Of these four, one came to
the session to complete the tests and the questionnaire and three
completed the same tasks over the phone. The other two attended
mainly to seek confirmation of their ways of applying skills to address
current life issues potentially serious enough to cause a relapse, as well

155



as complete tests and questionnaires. The absence of significance
between means of the tests administered at post sessions and the follow­
up is hypothesised to reflect the absence of substantial room for
improvement and also the robustness of the intervention beyond the
delivery of treatment. The third and fourth participants (Cathy and Dee)
commented that even though they did not need to have any sessions
during the course their availability assured them of help in mastering
concepts that could not be grasped solely online.

(b) Progress measured by the mood monitor and the Goldberg tests
The SOl and the HDRS were applied at the end of defined stages of the
intervention. The mood monitor and the Goldberg tests were completed
several times by the participants as they progressed through the course.
Scores on these two measures reflect the path of progress through the
course. Figure 2 presents a plot of mean mood monitor values for the six
completers at the nine different administrations throughout the course.

Figure 2. Graph of mean mood monitor scores over nine administrations

Mean mood monitor variations

8.0
~ 7.0
CDf 6.0
~ 5.0
~ 4.0
~ 3.0
-g 2.0
~ 1.0

0.0
2 3 4 5 6 7 8 9

Mood monitor administration

The graph shows a steady improvement in mood as participants advance
through the course. A t-test conducted on the means of the first
administration and the ninth administration indicated a significant
difference (t=-3.04, p<.01). The flattening between the fifth and the sixth

156



administrations is caused by two participants receiving adverse personal
news. However, both soon resumed their trend towards improvement.
The reversal for these two participants is highlighted in Figure 2. They are
depicted by a triangle and a star respectively (Participants 3 and 5). The
reversal of improvement demonstrates the influence of adverse life
events on depression while the resumption signifies that the individual
has addressed the setback and moved on.

Figure 3 depicts Goldberg scores for individual participants over the
course. A Goldberg score of less than 25 signifies no clinical depression.
The graph shows all participant scores at the seventh (last) administration
are below the cut off.

Figure 3. Goldberg scores for individual participants over the course.

Mean Goldberg Scores: Administrations 1-7

70
60
50

f 40
o
~ 30

20
10
o

1 2 3 4 5 6 7

--+- Participant 1
--II- Participant 2

Participant 3
Participant 4

--- Participant 5
---- Participant 6

Adm inistration

Note: participant 5 was at mild level when he began the course.

Figure 3 shows that reductions to below the cut off of 25 were achieved
by participants 2 (5th Goldberg administration), 3 (2nd Goldberg with a
temporary relapse at the 6th ) and 6 (3rd Goldberg). These occurred in the
first four weeks. The results are consistent with research by lIIardi and
Craighead (1994) which proposes that about two thirds of change in CST
for adult depression occurs within the first four weeks. The results also
appear to be consistent with the distribution of the learning curve, where

157



most of the learning occurs early and is followed by a plateau denoting
consolidation (for example Ritter & Schooler, 2002).

The following table draws data from the case studies to illustrate
significant changes in participant progress shown in Figure 2:

Table XVII. At 4 weeks into the intervention: Goldberg Scores, reporting
periods completed, key events in participant outlook

Participant Goldberg Reporting period Key events in first 4 weeks on
No depression level completed course

at 4 weeks (10=last)
1 - Aaron Mild 2 Overcame procrastination;

applied concept of looking at
things differently

2 - Bill Subclinical 9 Completed course three days
later

3 - Cathy Subclinical 9 Close to completing course
4 - Dee Mild 5 Grasped and applied concept of

looking at things differently
5 - Eric Subclinical * 10 Grasped and applied thought

challenging at home and work
in many different situations

6 - Fred Subclinical 9 Redefined his situation as less
threatening

* four weeks after delayed course commencement. Depression dropped
from mild to subclinical during the first four week period.

It appears that participants who advance more quickly through the course
show a higher rate of improvement. One explanation might be that they
are motivated enough to drive themselves to get well as quickly as
possible. Alternatively the more effort they put into the intervention the
quicker they achieve freedom from depression because the key

behavioural change skills are presented in the first few sessions.
Participant no 2 (Bill) is an example. He completed the module on
'looking at things differently' (the Epictetus principle) in the same day as
his initial interview and began applying the concept to his day-to-day
functioning. The table following, which shows the number of days taken
by each participant to complete the course, indicates that the four-week
mark tends to occur in the second half of course duration, with the

158



exception of the first participant. The table seems to support the idea that
early occurrence of change probably acts as one motivator for the
participants to persevere with the course.

Table XVIII. Days taken to complete the Internet course
Participant No of days
1 125

2 31

3 44

4 55

5* 54

6 36
Mean (SO) all 57.5 (34.4)
Mean (SO) (without outlier) 44.0 (10.6)

Note: Participant 5 married just after being tested and began the course
two months later. That is he devoted 54 of the 114 days since
enrolment to the course. His symptoms had largely abated by his
return but he wanted to do the course. He felt that as the novelty
and excitement of the marriage wore off and his mother-in-law
increased her dominance over the couple, depression might return
and he would need to have the skills taught by the course to
address the condition if it arose. (This reminds of the view by
Seligman (2006) that we should aim to maintain wellness rather
than solely cure disease).

If the outlier value is removed from consideration the mean length of time
taken by the five participants is 44 or about seven weeks and all
completed the course in less than two months.

(c) Follow-up assessments
In Table IX two comparison studies and the present study include follow­
up data. T-tests on the means of post-treatment and follow-up SOl
scores for each of the studies reveal that the difference is not significant.
Any gains made as a result of treatment have not changed by with the
passage of time; they have been maintained at the follow-up stage in all

159



studies. T-tests of the BOI means of the present study and the two
comparison studies reveal that the present study has a significantly lower
BOI score at follow-up (t=3.05 p<.05 for Shapiro et al. (2004); t=3.93
p<.01 for Jacobson et aI.1996)). Any interpretation should be tempered
because one of the comparison studies has a six-month term follow-up
and the other two including the present study have a three-month term.

Testing the difference with the t-test in proportions of compieters with
BOI<9 in the present study and in the comparison studies combined in a
group (by reworking data in Table IX) reveals that the difference is
statistically significant at the follow-up stage (t=2.68, p<.05). This leads
to the conclusion that the present study has a significantly higher
proportion of completers with BOI<9 at follow-up than the combined
comparison studies. As the case studies reveal, all completers and non­
completers fared well in their lives after completing the intervention. Their
depression lifted as they resolved key challenges in their lives such as
family problems, life issues and finding satisfying work.

(d) Conclusion
The question would seem to find some support from the results of this
study. Treatment gains, whether measured by mean differences or by
proportions of completers with subclinical symptoms, persisted for at least
3 months after completion of treatment. It needs to be stated clearly that
more research is required to establish the persistence of gains.

(iii) Research question 3
Is this intervention more likely to retain clients through to completion
compared to other interventions?

(a) Face-to-face depression intervention studies
Table VIII reveals 67% (6) of participants at the start of the program
completed all modules of the course and the face-to-face treatment. The
33% (3) non-completers did at least one module of the course. One of

160



the two female non-compieters improved to the extent that she felt that
she did not need to go on to complete the treatment. The other female
non-completer married just after completing the first course module and
felt that her life had changed for the better. The male non-completer
considered that a combination of contracting work he enjoyed and
applying thought challenging skills learnt in the course would provide him
with the means to manage his depression.

Comparisons in completion rates with other depression treatment studies
can only be approximated because every study is unique. Treatment,
provider and participant situation and experiences all vary. An
examination of completion rates of the comparison studies attempts to
place this study in context, as follows.

Table IX outlines the numbers of participants and completers for the
comparison (face-to-face) and the present studies. A Fisher Exact test
finds no significant differences between the ratios of completers to
participants in the comparison and the present studies (p> .05). Any
conclusion can only be tentative because of the small sample. Within this
limitation it may be tentatively concluded that the completion rate for the
intervention in the present stUdy is comparable to the completion rate for
face-to-face treatment. The results of the present study do not support
the proposition that the intervention is more likely to retain clients through
to completion compared to other interventions

(b) Internet or computer-assisted depression intervention studies
The results of the present study are compared with similar results from
other studies utilising online or computer-assisted treatment. The studies
are those analysed in Chapter 2 section (v). The following able presents
completion rates for the comparison studies and the present study.

161



Table XIX. Comparison of completion rates for Internet- or computer­
based CST depression treatment studies including the present study

Study Mean length of Intenders Completion
treatment (weeks) (Completers) rate

Andersson et al. (2004) 10 57 (36) 0.63
Christensen et al. (2004) 6 525 (414) 0.79
Clark et al. (2002b) 32 297 (220) 0.74
Marks et al. (2003) 9.2 55 (39) 0.71 *
(Cope program)
WriQht et al. (2005) 8 15 (12) 0.75
Total above studies 946 (721) 0.76
Present study 8.2 6 0.67

* total completion rate for all four programs in the Marks et al. (2003)
study

The table indicates that completion rates lie within a narrow band for all
studies. A Fisher Exact Test reveals that the difference between the
proportions of completers in the comparison study and the present
studies is not significant (p>.05). The present study is not different from
the other IT-based studies in terms of completion rates noting that it
differs in that the Internet delivery is integrated with face-to-face delivery.

A comparison with the Internet-based treatment studies in Table XIX
reveals that the present study has a high completion rate and the highest
rate for at least part-completion. The completion rate of the other
Internet-based studies combined (rate=0.71, n=683) was compared with
the completion rate in the present study (rate=.67, n= 6) by using the
normal approximation of the hypothesis test of the population proportion
(Wessa, 2006). The result reveals that the difference in completion rates
is not significant. That is the proportions of completers in the present
study and in the comparison Internet studies with the most completers
are similar.

162



(c) Conclusion
Research question 3 would seem to find some support from the results of
this study. The proportion of completers in the intervention is statistically
similar to the proportion of completers in face-to-face studies and other
online studies. However, the small number of participants in the present
study make this finding extremely tenuous.

(iv) Research question 4
What are the potential savings of this intervention compared to other
interventions?
There are four face-to-face comparison studies that include information
on the number of sessions provided to participants. The table below
outlines details of the number of sessions in the comparison studies and
the present study.

Table XX. Sessions provided to all participants and completers in face­
to-face studies combined and the present study

Comparison Present stUdy *
studies *

No of participants 272 9
No of sessions 3152.8 34.5
Mean sessions per participant 11.59 3.83
No of compIeters 209 6
Mean sessions per completer 15.09 5.75

* studies used: Shapiro et al. (1994), Cahill et al. (2003), Jacobson et al.
(1996) and Jacobson et al. (1991).

Fisher exact tests for the differences in the ratios of sessions to
completers and to participants for the combined studies and the present
study are significant (p<.01 for both tests). The raw data indicates that
the mean number of hours in the present study per participant (3.83) is
620/0 less than for the face-to-face studies.

163



The case studies reveal reductions in indirect treatment costs. Sy
undertaking the intervention at home participants do not have to take time
off work to attend the provider and also save travel and parking costs.

(a) Conclusion
Research Question 4 would seem to find some support from the results of
this study. The use of the intervention suggests that treatment costs may
be reduced in the order 062%.

(v) Note on interrelationships of different measures
The study utilises four different measures of depression. Two of these,
the SOl and the HORS, are applied in face-to-face sessions at defined
end points during the intervention namely pre-treatment, post-course,
post-sessions and follow-up. The mood monitor and the Goldberg tests
are completed by the participants at the end of particular course modules.
The question arises as to the homogeneity of measurements and the
causes underlying any apparent discrepancies in the measurement of the
same phenomenon, namely the participants' level of depression.

The interrelationships between the different measures employed in the
intervention to indicate levels of depression are highlighted in the table
following, which presents product-moment correlation coefficients for the
SOl with the other three measures at the different phases of the study.

PValue

Table XXI. Product moment correlation coefficients for SOl and other
measures at each study phase

Phase
Pre-treatment BOI - HORS Intend-to-treat 0.78 <.01

Completers 0.94 <.01
Pre-treatment BOI- Goldberg 1st administration 0.89 <.01
Pre-treatment BOI- Goldberg 2na administration 0.67 ns (.07)
Pre-treatment BOI- Goldberg SIn administration 0.53 ns
Pre-treatment BOI- mood monitor 1St administration -0.46 ns
Post-course BDI - Goldberg 7m administration (final) 0.55 ns
Post-course BOI - mood monitor 9ln administration -0.04 ns
Post-course BDI - HDRS 0.77 <.05
Post-sessions BOI - HDRS 0.42 ns
Follow-up BOI-HORS 0.44 ns

164



In the table the HORS and the SOl provide similar results at pre-treatment
and post-course. At the post-sessions and the follow-up stages the
distributions are no longer similar. The reasons might include that the
tests measure different aspects of day-to-day mood fluctuations for
individuals no longer clinically depressed, or that individuals might exhibit
too few symptoms to provide a valid measure of any remaining
depression on either or both scales.

The smoothness of progress through the intervention is illustrated in the
table following which lists Goldberg - product moment correlations of
progressive administrations.

Table XXII. Goldberg - product moment correlations of progressive
administrations
Administration r= p

1st and 2na 0.92 <.01
2na and 3ra 0.97 <.01
3ra and 4m 0.90 <.01
4tn and 5tn 0.81 <.05
5m and 6m 0.94 <.01

6m and 7m 0.96 <.01

The first administration of Goldberg, undertaken by participants at the
start of the course, provides scores with a distribution significantly close
to the pre-treatment SOl indicating that for this sample the two tests are
measuring statistically similar entities at this stage. As participants
improve in mood the relationship between Goldberg scores and pre­
treatment BOI scores weakens. This is to be expected because as
participants progress with treatment their pre-treatment scores are no
longer representative of their state of depression. The correlation
between the second Goldberg administration and the SOl is just short of
significance. Subsequent correlations move progressively further away
from significance.

165



The table shows that each administration is significantly correlated with
the one following. This indicates that progress throughout the course
averaged over all participants has been smooth, that is, without
significant discontinuities or statistically significant 'leaps'.

The compatibilities between mood monitor and Goldberg scores are
highlighted by the results of a Kolmogorov-Smirnov two-sample test. The
test produces a coefficient of 0.20 that is the difference in the distributions
of the two tests is not significant. The statistical similarity of the
distributions indicates the possibility of redundancy of one of the
measures and also reveals that mood monitoring is a robust measure
notwithstanding that it is subjectively derived. Individual correlations
between Goldberg and mood monitor scores at similar phases through
the course reveal that the two measures are not related including at the
final administration. Both measures serve psychologically important
functions. Completing the mood monitor requires the participant to focus
on day-to-day fluctuations and address their cause. Goldberg test scores
provide feedback on overall progress at certain stages throughout the
course and act as reinforcement of effort. Higher scores in SUbsequent
administrations of the Goldberg may underlie insufficient effort, the effect
on mood of a recent disruptive event and the possibility of saturation,
namely that a limit has been reached of possible gains through CST at
that time.

(vi) Case studies
The individual case studies outline a description of participant
backgrounds and progress through the intervention. The description is
based on information provided through face-to-face treatment sessions,
emails and questionnaires completed by participants at the end of the
phases of the intervention namely the course, face-to-face sessions and
the three-month follow-up. The names used are aliases.

166



Participants' attractions to the intervention, relapses, significant
components in the intervention and resultant behaviour are summarised
in Table XXIII.

(vii) Completers
Below is an outline of key points in each case. A fuller description is in
Appendix IV.

(a) Aaron
Aaron was first seen on 9 September 05. A male Caucasian in his 40s,
he worked as a youth counsellor. He has had depression intermittently

since an unpleasant childhood. He saw his life at a standstill since his
thirties. Aaron had been on antidepressants for several years.

His score on the BDI (Beck, Steer, & Brown, 1996) was 27, just below the
'severe' cut-off. The HDRS (Hamilton, 1960) score of 26 indicated a
moderately severe level. His first Goldberg score was 43, denoting
moderately severe depression, and the mood monitor showed an

average of 5.7 (O=bad, 1O=good) with little variation over the three days
and the three daily occasions. Two relapses were experienced resulting
in the first two reporting periods taking 48 days to complete compared
with 67 days for the remaining six periods.

In a face-to-face session Aaron practised the Epictetus principle to
resolve 'road blocks'. Two further sessions helped him consolidate his
knowledge before he submitted the fifth module. He notified his Goldberg
score of 22, denoting mild to moderate depression, down from 38 two

months earlier.

In his session at the end of the course Aaron revealed that he had begun
to achieve a progressively deeper understanding into the mechanisms of
addressing his depression from about the sixth reporting period onwards.
During this period he learnt negative thought challenging using

167



Aristotelian logic and Socratic questioning. He gained expertise in
applying the Epictetus principle. He considered his key tools were
negative thought challenging and looking at things differently. Goldberg
scores reduced to 7 at the end of the course.

Testing at the end of the course showed a SOl score of 0 and the
Hamilton score of 3, that is, no clinical depression. At the three-month
follow-up the SOl and Hamilton scores of 6 and 5 respectively indicated
no clinical depression as for the previous two occasions. Aaron 'it's been
a long time since I had no depression. It's a great way to treat'.

Key factors of participant experience
History: long-term depression, ineffectiveness of antidepressants
Attractions to online study: in charge of own treatment, face-to-face
sessions, blend with life
Relapse: yes and recovered
Significant component: Epictetus principle (looking at things in different
ways), negative thought challenging
Behaviour change: lifted depression and began University course

(b) Bill
The second participant, Sill, a graduate executive shift worker in the air
transport industry, was first seen on 9 October 05. He was depressed
and anxious. He had been seeing his GP for depression and had been
taking antidepressants. His capacity to resolve the problems amassed in
his personal life had not improved. He felt that his depression had led to
deterioration in his work performance, which might cause a disaster.
His wife of Asian background made little effort to acculturate and was
depressed. The couple's small boy was about the only joy for him. Sill
saw the resolution of his family problems as key to his wellbeing and to
improving work performance. His aim in treatment was to clear his mind
to reach a decision about his future in his family. Thoughts of breaking
up were clouded with intense gUilt feelings of abandoning his child.

168



Bill was attracted to the concept of the Internet training course because
he could progress through the modules at home between shifts. Live
sessions attracted him because he could discuss his difficulties with a
professional with whom he had established a relationship through the
Internet course. He was an experienced computer and Internet user.

Bill's initial BDI score was 28, at the cut-off between moderate and severe
depression (Beck, Steer, & Brown, 1996). The Hamilton test score of 29
indicated moderate severe depression (Hamilton, 1960). Bill's first
Goldberg and mood monitor scores were 60 (severe depression) and 5.9.

Bill took 31 days to complete the course and he attended six sessions. In
his second face-to-face session he presented in a better mood compared
when first seen. He found that he had little difficulty in understanding the
content of the modules and especially in putting the Epictetus concept
(looking at things differently) to work in his day-to-day life. He used the
first face-to-face session to help him clarify options about his marriage.
By the end of the session Bill had prepared a list of what he saw as
issues in the marriage which he and his wife needed to resolve.

Changes in Bill's Goldberg scores mirrored his perception of
improvement. In the first eight days his score fell from 60 to 45,
representing a category change from severe to moderate severe. It was
maintained at 45 during the next seven days. A further three days later
the score reduced to 27, at the low end of the moderate severe category.

Bill's end course assessment showed a score of 4 on the BDI and 2 on
the Hamilton (score 2). Depression symptoms were few and at a
subclinical level. The three-month follow up scores were 5 for the BDI
score and 4 for the Hamilton. No depression was noted.

Key factors of participant experience
History: depression due to continuing marital problems, ineffectiveness
of antidepressants

169



Attractions to online study: in charge of own treatment, face-to-face
sessions, treatment blends with life
Relapse: none
Significant component: Epictetus (looking at things in different ways)
Behaviour change: problem-solved marital issues, work performance
improved.

(c) Cathy
Cathy was an Asian woman from a well-to-do family in her mid 20s. She
was a graduate undertaking a master's research program. Cathy
became lonely in Sydney. A few months into her studies she became
close with a student from another Asian country and she fell pregnant.
She had an abortion which made her feel guilty. She did not reveal the
abortion to anyone else other than her partner. He did not give her any
emotional support. He was a gambler. She began to gamble herself and
soon became addicted. He often borrowed money from Cathy to feed his
habit, which he did not repay. Her funds dwindled, her depression
intensified and her studies began to suffer. Her partner was associated
with an Asian criminal gang operating at the Casino. He was arrested,
charged with fraud, convicted and imprisoned. Cathy was arrested
though later released. Her depression intensified as did her fears of
deportation.

Her first session was held on 2 October 05. SDI and Hamilton scores
were both 28. She was attracted to the Internet because she was an
experienced Internet user, liked information technology and could
integrate treatment with her day-to-day activities. She preferred to
undertake activities on her own though she appreciated the 'insurance'
that the availability of a live treatment professional provided her. Her first
Goldberg score was 44 (moderate severe) and her mood monitor
averaged at 4.7. Her second Goldberg was 25, just within the mild cut­
off. In the same period her mood monitor scores rose from an average of
4.7 to 6.1 further indicating a marked improvement. Cathy took 44 days to
complete the course. Apart from the initial and the after-course

170



assessment sessions, she attended one other. Her BDI and Hamilton
scores were at subclinical levels. In her post-course session on 22
November 05 held by phone because she was working long hours, she
gained a score of 5 in the BDI and in the Hamilton, both within the
subclinical level.

Cathy's improvement appeared as she began to come to terms with the
Epictetus principle though a temporary relapse did follow. Her three­
month follow-up showed a BDI score of 6 and a Hamilton score of 3.
Neither showed clinical depression. Cathy said that her life had been
more settled and she had been ruminating less about her distressing
recent past.

Key factors of participant experience
History: depression due to loneliness, damaging relationship, association
with criminal activity

Attractions to online study: Internet user, liking for computing, in
charge of own treatment, face-to-face sessions available if needed,
treatment blends with life
Relapse: recovered
Significant component: Epictetus, Aristotelian logic challenging
Behaviour change: resolved issues, continued with University course

(d) Dee
The fourth participant, Dee, a counsellor in a charity, was in her early 30s.
She had been married and had three children, now aged 8, 10 and 12.
Her husband was a violent drug addict who frequently abused his family
and assaulted her in front of her children. One day he attacked her with a
knife and inflicted serious injuries on her. Dee was markedly traumatised.
At a moment of weakness she allowed her parents to become the
children's foster parents. She had been placed on medication by her GP
but she realised that this made her aggressive and restless, and also
confused her thinking. She wanted help to lift her depression so that she
could make the appropriate life decisions.

171



Dee reasoned that the Internet course would enable her to blend
treatment with her life and not interfere with her work or study. The
opportunity for live sessions was seen as a safeguard. She did not
envisage any difficulties because she was familiar with CBT.

Dee's initial BDI and mood monitor scores were both 26 indicating
moderate depression. Her first Goldberg score, reported on 19 October
OS, was 46 and denoted moderate depression. The mood monitor
average was 4.6.

Dee completed the course in 55 days. The Goldberg score reduced
steadily from 46 during the course with no apparent relapses. The mood
monitor scores mirrored her improvement. In her post-course session she
scored 7 for the BDI and 6 for the Hamilton indicating no clinical
depression. Dee did not seek any sessions during the course. In her
three-month fOllow-up by phone the BOI score was 5 and Hamilton was 6.
No clinical depression was indicated.

Key factors of participant experience
History: depression following trauma, antidepressants ineffective,
loneliness without children, relapses
Attractions to online study: In charge of own treatment, face-to-face
sessions available if needed, treatment blends with life
Relapse: none during course
Significant component: problem-solVing
Behaviour change: problem-solved key issue of reuniting family

172



(e) Eric
The fifth participant, Eric, was in his mid 20s. He was of Middle East
background and his parents were devoutly religious. A charge of assault
led Eric to undertake a psychological assessment as part of his case in
court. He was diagnosed with depression, which had its roots in the ever
increasing conflict between himself and his parents over his Westernised
lifestyle and also in an earlier truculent relationship. His depression
appeared to be markedly worsened after he began to take steroids as
part of body building. The court placed him on a bond provided he attend
treatment for his condition. His GP agreed that he should receive
treatment from the writer. He attended his first session on 19 October 05.
Eric scored 24 for the SOl and 21 for Hamilton, indicating moderate and
moderate-severe depression respectively.

Eric was computer literate. He liked the idea of taking charge of his own
treatment and undertaking the course at his own time and pace. The
Internet course enabled him to blend his treatment with minimal
disruption to his lifestyle. He had given up steroids immediately after he
offended though he kept up with his gym.

Eric married after his court appearance. Eric wanted to complete the
treatment because he felt that his underlying depression, currently
abated, was likely to return.

Eric began completed the first module on 22 December 05, just after he
returned from his honeymoon. His initial Goldberg score was 10,
signifying mild depression. The mood monitor average was 8.3. His
Goldberg score reduced to 6 during the course signifying a change in
category from mild depression to subclinical. His mood monitor scores
varied between 8 and 9. Eric progressed well with the course. He took
54 days to complete all requirements. The end course SOl and Hamilton
(3 and 4 respectively) showed no clinical depression.

173



Eric's fOllow-up assessment revealed SOl and Hamilton scores of 2 each,
showing no clinical depression.

Key factors of participant experience
History: depression due to culture clash, relationship break up, worsened
by steroid use leading to offending
Attractions to online study: in charge of own treatment, IT experience,
treatment blends with life
Relapse: none during course
Significant component: logic challenging
Behaviour change: able to manage ongoing conflict with mother-in-law
without aggression.

(f) Fred
Fred, an Iranian on a visitor's visa was in his mid 30s. He had a good
command of English. He was a Christian and had suffered frequent
persecution from the government and the wider community in Iran. He
trained as a photographer and worked in this field whenever the
opportunity arose. In Sydney he found the freedom that he wanted and
began the procedure of applying for resident status in Australia. He was
rejected at different levels. He engaged an immigration law specialist
who was preparing an appeal directly to the minister. The guilt he felt
over being a burden to his brother and family for the last few years, his
loneliness and feelings of being an alien in a society that may reject him
led him to depression.

Fred attended his first session on 15 December 05. The concept of
pursuing his own treatment through the Internet appealed to him. He did
not have a computer but know how to access the Internet. He considered
the intervention as an opportunity of improving his online skills. He
arranged undertake the course on the computer in the writer's reception
office. Fred was briefed on the Internet course and quickly developed the
necessary access skills. Fred's initial SOl score of 26 and Hamilton
score of 23 revealed his depression was at a moderate level.

174



Fred began the course on 15 December 05. His first Goldberg score,
reported three days later, was 34. This denoted moderate depression.
The mood monitor average was 3.3 ranging from 3 in the morning and
the evening, to 4 at midday. The low level of the scores was a reflection
of his overriding loneliness and anxiety over his circumstances.

Apart from the introductory session Fred did not need additional
treatment sessions. His progress was steady. By the end of the course
36 days later Fred's Goldberg score reduced from 34 to 12, signifying
mild depression. His mood monitor showed parallel improvement, from
an average of 3.3 to between 7 and 8. Given that there was no
abatement in the pressures confronting him and the likelihood of a forced
return to Iran was ever present, the result was accepted by Fred and the
writer as the best possible under the circumstances. His post three­
month follow up assessment resulted in a BDI score of 5 and Hamilton
score of 5. No clinical depression was indicated.

Key factors of participant experience
History: depression due to possible forced return to Iran with dire
consequences and loneliness
Attractions to online study: pursuit of own treatment through the
Internet, face-to-face sessions available if needed, treatment blends with
life
Relapse: none during course
Significant component: Epictetus
Behaviour change: saw his predicament in a different light and decided
to problem-solve.

(viii) Non completers
(a) Glenda
Glenda was in her late 20s and worked as a communications professional
in a large organisation. She had recently entered into a relationship to

175



which she brought her four-year old daughter from her former marriage.
Her partner was also a professional in a similar industry. Glenda and her
partner were experiencing difficulties making their relationship work.
She felt she was being punished for bringing another man's child into the
relationship. Stressors were building up for Glenda with led to
depression.

Glenda attended the introductory session on 24 October 05. She scored
20 on the BDI, the lowest score at the moderate depression level. Her
score on the Hamilton was 17, the highest score on the mild-moderate
category. Four days later she forwarded a Goldberg score of 26, just
within the mild to moderate category. Her mood monitor average was
6.2, rising steadily from 5.7 in the morning, to 6.3 at midday and ending
the day at 6.7. A rapid improvement followed by the next assignment.
She advised that she had completed the course. She had not forwarded
her scores or assignments because she felt she gained enough
knowledge to resolve her depression. She and her partner worked
through some of the modules together and resolved some of their more
pressing issues. Her distress had abated and she expressed her
satisfaction at having the course available on the Internet case she ever
needed to address a relapse.

Key factors of participant experience
History: depression due to family discord involving her current partner
and child from a previous marriage
Attractions to online study: treatment blends with life situation and not
having to devote time away from her family
Relapse: none during time on course
Significant component: problem-solving
Behaviour change: Negotiated changes with her partner, which resolved
her depression.

176



(b) Harry
Harry, a former armed forces careerist, was in his mid 50s. He had
accumulated some negative experiences which were often triggered in
his mind and made him depressed and anxious because they came close
to claiming some of his mates' lives. Harry experienced difficulties in
adjusting to civilian life and work.

Harry attended his first session on 29 September 05. He was a recent
but rapid learner of the Internet. Harry felt a little reticent in undertaking
the depression Internet course on his own with email help and the
occasional remedial session. He opted for weekly sessions in parallel
with undertaking the program. His armed forces career made him feel
more comfortable following a set structure.

Harry's BDI score was 27 (moderately depressed). The Hamilton score
of 36 was at the moderately-severe category. His first mood monitor
average was 4.3 and showed no variation throughout the day. His first
Goldberg score was 62 denoting severe depression.

Harry completed the first four modules and attended his fifth and last
session on 20 November 05. He subsequently kept in touch. On his last
module he reported a mood monitor average of 5.9 with a peak in the
evenings of 6.7 denoting that he was able to break the depression barrier
at least for part of the day. His Goldberg had reduced to 45 placing his
depression in the moderately severe grouping. Harry advised that he had
read the remaining modules. He felt that he gained enough knowledge
and wanted to put it into good use to lift his depression. He challenged
negative thoughts supporting his nemesis that he could not be gainfully
employed in civilian life. He managed to gain work as a contractor in an
industry which valued and drew on his armed forces skills.

Key factors of participant experience
History: depression due to adjustment to civilian life after a career in the
defence forces

177



Attractions to online study: technical challenge in pursuing own
treatment through the Internet, face-to-face sessions in parallel, treatment
blends with life situation
Relapse: none during time on course
Significant component: problem-solving
Behaviour change: found satisfying work, which lifted his depression

(c) lisa
lisa, was in her mid thirties. She had a difficult life. Her father left the
family shortly after her birth and she was brought up initially by her
maternal grandparents. Her mother remarried but the stepfather took
little interest in his wife's children. In the mean time her mother pursued a
number of extramarital affairs and the marriage eventually broke up. lisa
went to live with an older cousin initially and she was then farmed out to
other relatives. She changed schools several times because of her
frequent moves between her mother and relatives. She did not manage
to attend school beyond Year 10. She married at the age of 18 to find a
sense of stability in her life and had four sons in the next six years. Her
marriage destabilised as both partners pursued extramarital affairs and
finally broke up. While she was trying to pick up her life she was raped
by a group of men in her own home, who had been working in a nearby
building site. lisa completed TAFE training and gained work which gave
her a reasonable income. Her depression however continued to distress
her and would arise whenever she was reminded of her sexual assault
and her childhood abandonment. She came to the attention of the
authorities over a fraud charge and her solicitor advised her to see the
writer for a presentencing report. The writer diagnosed depression and
remnants of trauma symptoms. Her solicitor recommended that she
attend for treatment and her GP agreed.

lisa attended on 27 October 05. She was attracted to the Internet course
because she lived in a country two hours away from Sydney. The option
of attending sessions as required appealed to her. However she felt that
it was unlikely that she would be taking advantage of their availability,

178



preferring email contact if she had the need to do so. Moreover she had
formed a relationship with a man and they planned a wedding in the next
few weeks.

lisa scored 26 on the SOl which placed her at the high end of the
moderate category and 28 on the Hamilton, which was within the
moderate-severe grouping. She completed the first two modules. Her
Goldberg score was 72 signifying severe depression. Her mood monitor
average ranged between 2.2 and 2.7, the lowest of all participants.

lisa did not complete any more modules. She advised that her mood had
lifted considerably following her marriage. She had downloaded the
Internet course into her computer so that she could study it whenever she
needed help.

Key factors of participant experience
History: depression due to earlier trauma and life difficulties since
childhood.
Attractions to online study: no travel hence no drain on time;
availability of face-to-face sessions if needed.

Relapse: none during time on course
Significant component: none advised
Behaviour change: changed life circumstances

(ix) Summary of key participant experience
The key factors emerging from participants' experiences with the online
course and the sessions are summarised to provide an overview of the
key areas of significance in the intervention. The table following provides
the summary.

179



Table XXIII - Key factors of participant experience
Completers

History of Attractions to Relapse Significant Behaviour
depression intervention component{s) change

Aaron long-term, in charge, recovered Epictetus, lifted
antidepressants sessions, negative depression,
ineffective blend with thought began uni

life challenging course
Bill marital in charge, none Epictetus problem-

problems, sessions, solved
antidepressants blend with marital
ineffective life issues,

work
performanc
e improved,
depression
lifted

Cathy depression due in charge, recovered Epictetus, resolved
to loneliness, Internet user, logic issues,
damaging likes IT, challenging continued
relationsh ip, sessions, with
association with blend with University
criminal activity life course,

depression
lifted

Dee depression after in charge, none problem- problem-
trauma, sessions, solving solved key
antidepressants blend with issue of
ineffective, life reuniting
loneliness family,
without children, depression
relapses lifted

Eric depression due in charge, IT none logic thought able to
to culture clash, experience, challenging manage
relationship blend with ongoing
break up, life conflict with
worsened by mother-in-
steroid use law without
leading to aggression
offending or relapsing

into
depression

Fred depression due pursuit of none Epictetus redefined
to possible own predicamen
forced return to treatment t in a
Iran with dire through the different
consequences, Internet, light and
loneliness sessions, decided to

blend with problem-
life solve,

depression
lifted

180



Non completers
Glenda depression blend with none problem- Negotiated

due to family life, not solving changes with
discord due to having to her partner,
current devote time which resolved
partner and away from her
child from a her family depression.
previous
marriage

Harry depression technical none problem- Found
due to challenge in solving satisfying work,
adjustment to pursuing which lifted his
civilian life treatment depression
after a career through the
in the defence Internet,
forces parallel

sessions,
blend with
life

lisa depression no travel none none Depression
due to earlier involved; lifted due to
trauma and sessions if changed life
life difficulties needed. circumstances
since
childhood

Summary History of Attractions Relapse Significant Behaviour
depression to component(s) change

intervention
Completer life situations blend with relapsed Epictetus (4), Solved
s and Non life (6), but problem- underlying
completers sessions recovered solving (3), life issue,

(6), in (2) logic depression
charge of no relapse challenging lifted
treatment (7) (s), none (1)
(5),
technical
and IT (3),

The table reveals that depression for all participants, completers and non­
completers, stems from a range of life situations, mainly relationship
problems. The attractions to undergoing the intervention include the
ability to blend treatment with life's demands, the availability of face-to­
face sessions if needed, being in charge of one's own treatment and
using the Internet particularly for those who have an affinity with IT and
this medium of communication. The significant elements in the course,
namely those leading to a marked insight into one's depression, include
the Epictetus principle, logical challenging and problem-solving. All

181



completers succeeded in addressing their depression by taking steps to
resolve the issues confronting them or by looking at their problem in a
different light.

(x) The contribution of the case studies to the research questions
The case studies provide information on the research questions. The first
research question, namely "Is an intervention which relies on an online
course to train the client in cognitive behavioural techniques and on face­
to-face sessions to practise the techniques at least as effective as face­
to-face counselling in treating clients with mild to moderate depression?"
is partly answered. All six completers found the intervention valuable and
succeeded in reducing their depression to subclinical levels. Those adept
at self-learning were well able to progress on their own. The three non
completers used the skills in the course, without completing assessments
or other requirements to reduce their depression. Most had downloaded
the site to their computer for ready availability in the future. None had
undergone face-to-face treatment previously so that the intervention
could not be compared with this form for effectiveness.

The completers demonstrated that by incorporating the skills learnt in the
online course, such as looking at things differently and challenging
negative thoughts, in their day-to-day functioning they were able to
maintain a depression-free outlook beyond the completion of the course.
The third question cannot be answered fUlly. The intervention did retain
six participants to the conclusion because they saw an improvement
developing as they progressed through the course. Two non completers
gave up the course because they were able to benefit from the presented
material.

The fourth question can be demonstrated through its ability to help
several of the participants who had been on antidepressants to gain
freedom from this debilitating condition. Indefinite continuation of
antidepressants without resolution of depression would reach the stage

182



where their total cost would exceed the cost of any intervention which
proved to be successful.

(xi) Evaluation of contents and presentation of online course
The evaluation is undertaken with the aid of results from the
questionnaires. It needs to be noted that the comments are essentially
client-therapist feedback and this aspect would affect the veracity of the
comments. Hence care needs to be taken with any interpretation of the
comments. Statistics summarising participants' views on the intervention
as provided on the questionnaire were presented in the discussion on
Research Question 1. The following table outlines mean responses to
items in the questionnaire on contents and presentation.

Table XXIV Post module questionnaire on contents and presentation ­
mean scores

No Item Mean
1 The text in the module was easy to understand 5.00
2 The sessions had too much material for my level of concentration 1.50
3 The sessions had too little material to keep my interest 1.20
4 The material was boring 1.20
5 I enjoyed the way the material was presented in the module 4.80

Reflecting on the pictures on the right side of the pages helped me feel
6 better 4.17
7 The humorous pictures were enjoyable 4.40
8 I always visited the sites with more information 4.00
9 The sites with more information were useful 4.50
10 I always visited the qames and puzzles sites 2.50
11 The games and puzzles were a good way of relaxing after each session 2.50
12 I ignored many of the sites on the right 3.67

Completing Goldberg was a good way to let me know how I was
13 progressing 4.60
14 The exercises helped me to apply the skills I learnt in the sessions 3.67
15 The exercises took too long to do 1.20
16 The exercises were not long enough 3.67
17 Emails enabled me to communicate adequately with my psychologist 5.00
18 Even though my psychologist wasn't with me I felt that I had his support 5.00
19 I came close to giving up on the module 1.50
20 Looking back on the module, I am glad I did it 5.00

Participants chose responses that ranged from 5 implying 'strong
agreement' with the statement through 4 'agreement', 3 'not sure', 2

183



'disagreement' and 1 'strong disagreement'. The table yields the following
results.

(a) Quality of communication
Quality of communication (items 1-7) was defined as readability, interest
and manageability of material presented in the modules. There was full
agreement about the readability of the material. All completers rated
readability at the maximum mark of 5. The result becomes more
significant when it is considered that three of the completers are of a non
English speaking background and speak another language at home. The
sufficiency of material in each module and its interest rated almost as
highly. Three completers scored 1 (strong disagreement) for the item
stating that the modules had too much material for the participant level of
concentration while the other three scored 2. Similarly five completers
rated at the most favourable score the items concerning boredom,
insufficiency of the material to maintain interest and enjoyment. One
completer rated the items at one point less than the best score. The
opportunities for reflection and humour in the course were similarly rated
highly. No completer mentioned in the assessment session any difficulties
or suggestions about the quality. On the basis of completer comments
the course material appears easy to read, able to maintain interest and
comes in manageable pieces.

(b) Enrichment material
The enrichment material (items 8-12) consists of sites with further
information on the topics covered in the course modules, for example on
Epictetus, CST and Aristotle. Five completers scored four or more on
visiting the sites always and one completer scored two. During his
session this completer (Aaron) remarked that he visited some sites and
read until he felt the material became irrelevant. All compieters scored
the usefulness of the information on the sites at 5 or 4. However five
completers ignored many of the sites. During their post course
assessments no particular patterns emerged - it appeared that whether
or not a completer clicked on the site depended on interest, perceived

184



relevance, mood at the time, tiredness, time of day and other time
pressures.

(c) Exercises
The course contains several exercises for completion by the participant to
be forwarded to the writer. The exercises (items 14-16) cover the
application of key principles in CST. Five completers disagreed at the
strongest level (score of 1) that the exercises took too long to do. One
completer gave the question a score of 4. Whether the exercises were
long enough polarised the comp/eters. Four agreed strongly that the
exercises were not long enough and two disagreed strongly. One of the
two who disagreed (Eric) commented that he devoted more time to
practising the skills in vivo because he was more practically oriented.
The other (Fred), who had some difficulty with written English, felt that
they were the right length for him. The capacity of the exercises to help
participants apply the skills learnt in the modules polarised the
completers. Three agreed strongly (score 5) and one agreed at score 4.
Two disagreed scoring the item at 1 and 2 respectively. Of these one
was an experienced CST user (Dee) and the other (Eric) felt that he
gained more by applying the skills to his day-to-day functioning. The
results indicate that length, usefulness and time demands of the
exercises are largely determined by individual perceptions.

(d) Live support
All completers rated at 5 their agreement that they perceived that they
had their treatment provider's support (items 17-18) even though he was
physically not present. Communication by email was apparently
considered to be adequate. Email support entailed sending and receiving
assignments and depression test assessments, seeking advice on 'stuck'
points and arranging live sessions. The result supports the acceptability
of online treatment and live support through asynchronous
communication.

185



(e) Motivation to complete
All completers on items 13, 19 and 20 disagreed (scores 1 and 2) that
they came close to giving up the course and all strongly agreed (score 5)
that they were glad to have completed it. They agreed at score levels 4
and 5 that completing Goldberg provided a good measure of progress.
The conclusions are that the course is perceived to be sufficiently self­
motivating, the frequent feedback of progress is appreciated and
completion is worthwhile.

(f) Evaluation of questionnaire on the course content
No significant shortcomings were highlighted in the analysis of results.
The course content appears to communicate well to participants and is
able to motivate them to complete its requirements. The enrichment
material appears at least sufficient if not too extensive. On the whole the
exercises appear to be acceptable though perhaps they could be
redesigned to introduce optional additional components and to strengthen
their relevance to day-to-day life. Frequent completion of the Goldberg
Depression Scale is considered important as feedback to effort. Email
communication appears sufficient to assure the participants of access to
their treating professional. The next table presents data on behaviour
changes resulting from the course.

186



Table XXV. Questionnaire: Behaviour changes resulting from course­
mean Likert scores (0-5) at post-course, post-sessions and follow-up

Post Post Follow-
No Item course sessions up
1 Avoiding isolation 4.5 4.5 4.25
2 Exercising 4 4 3.8
3 Getting up at usual time 4.6 4.4 4.2
4 Going on with daily activities 4.75 4.75 4.75
5 Breathing relaxation 4.25 4.75 4.25
6 PMR 3 2.2 2.2
7 Basic meditation 3.5 2.6 2.8
8 Mini meditation 4 4.75 4.6
9 Relaxation activities 4.75 5 4.5
10 Taking supplements 4.5 4 4.2
11 Following a good diet 4.75 4.5 4.5

Identifying negative thoughts causing a
12 downward mood 4.8 4.8 4.6

Challenging negative thoughts with 'where is my
13 evidence' 5 5 5

Challenging negative thoughts using Aristotelian
14 logic 4.5 4.2 4.2
15 Persevering with thought challenging 4.5 4.75 4.75

Note: the table makes no distinctions between the sub-groups of clients
by face-to-face sessions because only two participants (Aaron and Bill)
took post-course sessions. Those who took in-course sessions took them
at different stages of the course. An analysis of the questionnaire results
follows.

(xii) Transfer of skills
The questionnaire sought information on the extent to which skills learnt
in the course were transferred to life situations and whether the transfer
lasted beyond the course itself. The questionnaire tapped into three
broad areas - stress management, coping with life and CBT skills. An
analysis of responses follows.
(a) Stress management
Breathing relaxation appeared to be widely adopted. Three completers
reported ratings of 5 (the procedure is done as often as needed) from
three completers and 4 (more often than not) from the other three. PMR
(Progressive Muscle Relaxation) (PMR) proved to be somewhat less
popular. Only one participant (Eric), a gym enthusiast, rated it at 5 and in

187



an email he remarked that it helped him in his muscle building. Two
participants (Aaron and Dee) used PMR 'less than half the time needed'
and 'hardly ever' respectively; hence neither saw much value to it. Their
views on interview were that breathing relaxation was similarly effective
and less complicated. The extent of use remained essentially similar in
the following administrations of the questionnaire. At the follow up
administration breathing relaxation was rated at 5 by three completers
and 4 by the other three. PMR use was polarised - three compieters
used it hardly ever while the other three more often than not. The post­
session questionnaire yielded compatible results.

At course end meditation was reported to be used by three completers as
needed and by one more often than not. By follow up three completers
reported using meditation more often than not. Mini meditation was rated
at least 4 by five completers and by one (Cathy) at 3 or at least half the
time she needed to use it. In her interview Cathy said that breathing
relaxation was enough for her and that she preferred to analyse her
thoughts as a way of reducing stress. At follow up all participants said
that they used mini meditation at least more often than not. It appears
that over the three months since ending the course, completers come to
incorporate mini meditation as an often used approach in managing
stress.

Five participants rated usage of relaxation activities (reading, humorous
videos) at 5 and one at 4. The ratings were maintained in the
assessment at the end of the sessions. At follow up all completers rated
usage at 5. The conclusion is that the course assists participants to
appreciate the need for and incorporate relaxation activities as legitimate
tools in their efforts against depression. Exercising received ratings
similar to relaxation activities. Five completers rated exercising at 40r 5
in all administrations of the questionnaire. The sixth (Cathy) said in her
interview that she realised that exercising is important. However the
pressures of her studies did not give her much opportunity to exercise
though she did go to the gym in her apartment block as time allowed.

188



(b) Coping with life
All completers, except one, report in all administrations that they continue
to get on with their daily activities as required. The exception, Aaron,
does so more often than not. The amplification given in his interview is
that he faces challenges in balancing work and study partiCUlarly in
planning his assignments to avoid last-minute rushes. All completers
avoid isolation and communicate with significant others as often as they
need to do so, or 'more often than not'. The finding remains stable
through all three administrations of the questionnaire. Getting out of bed
at the usual time is also robust in all administrations with the exception of
one completer (Aaron) at follow up. His interview reveals that because
he studies until late at night he has a little more sleep in the morning
provided he does not have to go to work. All completers report in all
administrations that they follow a good diet and take supplements at least
more often than not. The conclusion from this part of the questionnaire is
that the program succeeds in assisting participants to transfer knowledge
and skills learnt in the module to their day-to-day lives.

(c) CST skills
The questionnaire measures the frequency of use of four key approaches
in the application of CST to daily life, which concern the identification and
challenge of negative thoughts causing a downward mood. In a practical
sense a key measure of success of the course is the extent to which

completers incorporate these approaches in their functioning. Responses
show that the most often used approach is the Socratic question 'where is
my evidence'. All participants in all administrations of the questionnaire
rate this approach at 5 - it is used as often as needed. Identifying
negative thoughts causing a downward mood and persevering with
thought challenging until the mood changes are almost as highly rated as
Socratic questioning. In the post course administration four completers
rate these approaches at 5 (as often as needed) and two at 4 (more often
than not). In the post session and follow up five completers rate it at 5.
Eric rates both at 4 because, as he said in his interview, sometimes he

189



finds difficulty in identifying and challenging negative thinking in situations
where he and his mother-in-law exchange heated words. He reiterates
the difficulty he finds in looking at his mother-in-Iaw's views differently but
realises he needs to continue to try to do so. Follow up interviews reveal
that, as with Eric, a major plank in persevering with thought challenging is
the Epictetus principle. In regards to Aristotelian rules of logic five
compieters rate the item at 5 in the post course administration of the
questionnaire and four at the subsequent administrations. One (Eric)
rates the approach consistently at 3. Eric finds the Socratic and Epictetus
principles more readily applicable. In summary the analysis of this part
of the questionnaires, illuminated by the compieters' views, illustrate that
the principles have been successfully transferred from the course into
their day-to-day functioning.

(xiii) Evaluation of the efficacy of the two components of the
intervention
The issue arises whether the efficacy of each component of the
intervention, namely the online course and face-to-face treatment can be
evaluated. The results prOVided in answering the research questions
demonstrate the efficacy of the intervention as a whole. For example
Table X shows that the study achieved subclinical scores on the SDI for
all completers. In addition Table XXIII provides evidence of the success
of the intervention from the participants' own realities.

The intervention is an integration of two components - an Internet course
and face-to-face sessions. It is difficult to evaluate the effectiveness of
each component separately in the light of the reality that participants
arranged and attended sessions whilst they were undertaking the Internet
component. The picture becomes more blurred when it is considered that
there was email contact by the writer with all participants and that all were
seen before the commencement of the intervention. However given that
four completers did not seek sessions during the course it might be

190



assumed that the course alone was an efficient mode of treatment for
them. Participant views are analysed below.

(xiv) Participant views on the intervention
Participants completed items based on Likert scales with range of 1
(strongly disagree) to 5 (strongly agree) to assess their views on different
aspects of the Internet course. The instruments are presented in
Appendix (ii). Scores are drawn from several item scales to answer the
research question.

The following table outlines items and scores from the post course
questionnaire which indicate participant satisfaction. The first two items
draw average responses at the maximum level of 5 signifying high-level
participant satisfaction with the course.

Table XXVI. Post course questionnaire views on participant satisfaction ­
mean scores

Direct evidence on satisfaction
No Item Mean

19 Looking back on the course, I am glad I did it 5
20 Revisiting the course should help me avoid depression in the future 5

Supporting evidence on satisfaction
1 The text in the course was easy to understand 5

2 The sessions had too much material for my level of concentration 1.5

3 The sessions had too little material to keep my interest 1.2

4 The material was boring 1.2

5 The way the material was presented in the module was enjoyable 4.8

Although the presentation of CST differs between psychologists there are
common principles such as those which appear in the table below.
Participants completed scales on items evaluating their use of the skills
taught from 5 (as often as you need to do so) to 1 (hardly ever) as
follows:

191



Table XXVII. Post course questionnaire views on use of skills taught­
mean and range of scores
No Item Mean Range
112 Identifying negative thoughts causing a downward mood 4.7 4-5
113 Challenging negative thoughts with 'where is my evidence 5 5-5
114 Challenging negative thoughts using Aristotelian logic 3.7 3-5
115 Persevering with thought challenging until the mood improves 4.5 4-5

The response levels indicate that completers consider that they are able
to apply key CST principles effectively in their day-to-day functioning.
The same items administered three months after completion of the
sessions reveal little variation as the following table shows.

Table XXVIII. Post course and post three months response comparisons
on use of skills taught- mean scores

Mean
Mean post
post three

No Item course months
112 Identifying negative thoughts causing a downward mood 4.7 4.6
113 Challenging negative thoughts with 'where is my evidence 5.0 5
114 ChallenQinQ neQative thoUQhts usinQ Aristotelian IOQic 4.0 3.6

Persevering with thought challenging until the mood 4.7
115 improves 4.5

There is little change in items 112 and 113. Item 114 shows a reduction
of 0.4, from 4.0 to 3.6 while item 115 shows an increase of 0.2 from 4.5 to
4.7.

The analysis indicates that participants see the course as a legitimate
treatment because it provides training in key skills in the amelioration of
depression and also, combined with face-to-face treatment, facilitates the
embedding of the skills in individual day-to-day functioning.

The free written comments added by three participants to their
questionnaires and verbal comment by the other three highlight their
satisfaction with the intervention, as indicated in the following table:

192



Table XXIX. Free comments by participants on value of intervention
Participant Free comment

Aaron (written) Well structured & presented. A good tone achieved in the
writing.

Bill (written) The best thing I have ever done. My life has changed
dramatically since the beginning of my treatment but I believe I am much
better equipped to deal with issues in a more logical, stress free or
stress reduced way. I do feel confident that I can prevent depression.
Excellent site. Helped my progress exponentially. Will revisit site every
2-3 days to go over topics and keep fresh with techniques to help
overcome my depression.

Cathy (written) I liked the module, it's easy & straight forward and also had a
lot of teachings that I could relate too. I like its convenience as I can
access it whenever I want being online. The module made me think, &
the exercises were helpful as they made me think. I especially find the
thought-challenging exercises helpful.

Dee (verbal) Lifting of her depression enabled her to resolve key issue
concerning her reunification with her children

Eric (verbal) Learnt to apply a greater tolerance to events which formerly
provoked irritable responses

Fred (verbal) Saw his predicament about a negative response by the
government to his refugee visa application as less catastrophic

The discussion illustrates that the completers see the Internet course in
this study as a legitimate treatment vehicle. That is the course facilitates
the learning of appropriate skills in the treatment of depression arising
from a wide variety of causes. The skills that appear to be most effective
are those that emanate from the Epictetus principle, Socratic questioning
and Aristotelian logic.

The research questions are as follows:
1. Is an intervention which relies on an online course to train the client

in cognitive behavioural techniques and on face-to-face sessions to
practise the techniques at least as effective as face-to-face
counselling in treating clients with mild to moderate depression?

2. If a positive effect is realised from such an intervention, how
persistent might this effect be?

3. Is this intervention more likely to retain clients through to completion
compared to other interventions?

193



4. What are the potential savings of this intervention compared to other
interventions?

(xv) Summary
The findings presented in this chapter suggest that:
1. A partly Internet-based intervention could be at least as effective

as face-to-face counselling in treating clients with mild to moderate
depression;

2. Gains persisted three months after completion of treatment;
3. The proportion of completers in the intervention was statistically

similar to the proportion of completers in face-to-face studies or
other online studies;

4. Considerable savings measured in saved provider time compared
with face-to-face could be realised.

194



CHAPTER 5 - DISCUSSION

(i) Genesis of the study
The genesis of the study emanates from the debate emerging on the use
of the Internet for clinical psychological interventions. Studies by Clarke
et al. (2005) and Christensen et al. (2004) demonstrated the efficiency of
the Internet in the early treatment and prevention of depression. The
researchers saw a radical change in health systems in developed nations
with self-help and self-responsibility for evidence-based interventions
becoming a new component of healthcare delivery. Fenichel et al. (2005)
summarised and responded to the more frequently heard misinformation
and arguments against delivery of treatment through the Internet,
including that online therapy was impossible, it was delivered almost
entirely by email, it was devoid of the richness of human experience, it
could not address suicide prevention and crisis intervention and that
online and face-to-face treatment shared similar principles. Many
studies, for example those referred to in this study, demonstrated the
efficacy of the Internet as a treatment delivery vehicle in a wide range of
applications from weight loss to potential suicide. The richness of human
experience can be provided by the participant's imagination in the same
way that poetry is enjoyed. Self-learning is a well established principle
and is compatible with self-responsibility for one's treatment that is valued
today. The use of self-learning in the present study emanates from the
writer's experience in using programmed learning techniques to train
RAAF technical trainees (Jacmon, 1966).

(ii) Restatement of the problem
The study aimed to provide insight into the use of the Internet as a
component of face-to-face treatment. The study was originally
conceptualised with the Internet course as the first stage in treatment. In
the first session the client would be assessed and if found with mild to
moderate depression would be invited to participate in the course. If

195



agreeable the client would be briefed on accessing the course, working
through the modules, completing required activities and emailing results.
The client would be invited to arrange to attend for face-to-face individual
sessions if difficulties were encountered in the concepts covered by the
course. Following completion of the course the client would be tested to
assess the current level of depression and to map out treatment
strategies. Next the client would be offered face-to-face sessions during
which guidance would be given by the treatment provider to apply the
skills learnt in the course to daily functioning. Testing would be carried out
at the end of the sessions to assess the effectiveness of treatment in
reducing depression or to gauge the need for further attention.

The concept of session usage needs to be redefined. Each participant in
the present study had one pre-course session and a further session was
estimated to have been given in relation to commenting on emailed
assignments and comments. Four of the six participants attended
sessions during the course itself. The post-course session was used
mainly for assessment. Only two participants required further sessions
for treatment purposes. In one case a personal difficulty had occurred
during the course which the participant wanted to discuss. In the other
case the participant suffered a relapse. The implications of these
events are that it is not necessary to draw distinctions between the
course and the session phases. In future applications of Internet
interventions clients would be invited to arrange sessions as they feel
the need to do so.

Another part of the problem that should be restated in the light of
experience concerns the limitation of participants to those with mild to
moderate depression. SOl scores from only two participants
(completers and non-completers) lay in the lowest half of the moderate
category. A completer with a pre-treatment SOl score of 28 (the upper
cut-off of the moderate category) scored in the severely depressed
category in the first administration of the Goldberg after he commenced
the course indicating that he may have been more than moderately

196



depressed. Five of the studies in Table IX had a mean BOI score lower
than that in the present study. Research is accumulating on the
effectiveness of CBT for severe depression, for example Winbrow
(2005) in discussing the NICE gUidelines states that CST is the
treatment of choice for moderate, severe and treatment resistant
depression particularly for young people. Severely depressed clients
presenting themselves for treatment in the future would not be
precluded from the course. They would be advised that research at this
stage is not clear that CBT is the best method available. If they choose
to undertake the course they would need to monitor their progress
closely and in the absence of any improvement they should be prepared
to seek alternative treatment.

In the study successful participants are deemed to be those who achieve
a subclinical level of depression on the BOI, namely below 9. The results
showed that all six completers met this criterion. Of the other three
participants, a woman who scored at the lower cut-off for moderate
depression (the lowest score in the study), left the study after completing
one module of the course. She advised that she became better qUickly
and was happy to learn the material on the website without submitting the
homework. Another participant, with initial BOI score just below the upper
cut-off for moderate depression, left the course after the third Goldberg
administration. He advised that the modules that he completed gave him
certain skills and made him realise that he needed to take charge of
himself by addressing the negativities that occupied his mind and
pursuing challenging activities. He was a former defence force specialist
and was able to set himself up as a consultant in his field of expertise.
He later advised that he had secured several contracts to do training and
that he was very busy. Importantly depression no longer worried him.
The third participant, a woman, completed two modules up to the second
Goldberg administration before her marriage. She reported that marriage
had changed her life and that she was no longer depressed. The
conclusion emerging from consideration of the non completers is that
they have been in fact successful in defeating their depression. One did

197



so early in the course, another used advice in the course to map out his
own therapy and in the third a change in life circumstances led to the
desired result. In the real-life application of the course any participant
who improves to the extent that depression is no longer at clinical levels
at any module would be considered a success of the intervention.

(iii) Major findings and their relevance to the research questions
Hypotheses were evaluated on the basis of comparison of five studies
similar to the present study except that they involved entirely face-to-face
depression treatment (Table VIII) and where appropriate the results of the
questionnaires. In all studies participants were treated with CBT and their
level of symptom distress was measured by the BDl. Comparisons were
made with studies offering IT-based treatment only (Table VIII).

(a) First research question
Whether an intervention which relies on an online course to train the
client in cognitive behavioural techniques and on face-to-face sessions to
practise the techniques at least as effective as face-to-face counselling in
treating clients with mild to moderate depression was tested by the
application of t-tests to the means of scores of participants in the present
study in the various phases (pre-treatment, post-course, post-treatment
and follow-up) compared with those in the face-to-face studies with
similar score categories. The first research question would seem to find
some support from the results of this study.

Online treatment with or without a face-to-face component has been
demonstrated to be appropriate for disorders including panic attacks and
Generalized Anxiety Disorder (Westen & Morrisson, 2001), weight loss
(Tate & et aI., 2001) and eating disorders (Winzelberg 2000). Given that
the studies for example in Table XVII demonstrate that CBT lends itself to
Internet presentation and that CBT is an evidence-based treatment, there
is an argument for further research to assess the merits of expansion of
online treatment. With the limitations of this study, further research is

198



merited to replicate this work, with tighter controls over the modalities and
the mix of modalities used and evaluating the impact of these with control
groups. The present study tentatively demonstrates that it is not
impossible for a provider to develop an online course embodying his or
her treatment approach.

A factor worth noting is that an intervention which integrates course and
sessions is essentially 'fail safe'. Where the client cannot cope with the
course because of low concentration or the need for maximum personal
support, he or she can simply discuss the issue with the provider and
revert to entire or increased face-to-face treatment. In short, the client is
wholly in charge of how the treatment is to be delivered. It is
hypothesised that a large proportion of clients would be motivated to try
the integrated treatment at least because of cost savings.

(b) Second research question
The second research question, the persistence of any positive effect
realised from such an intervention, would seem to find some support from
the results of this study. There is support from the literature for this
finding, for example Andersson et al. (2005), Christensen et al. (2004),
Clark et al. (2002b) and Marks et al. (2003).

It might be argued that the period of three months might not be lengthy
enough to demonstrate robustness of the intervention. However the
purpose of the intervention is to train participants in skills to address
downturns in their moods whenever they occur. The individual who did
not gain skills adequately during the intervention is unlikely to apply them
effectively in three months time or later. On the other hand the individual
who incorporates the skills in his or her daily functioning as they are
learnt and applies them as needed is likely to score at subclinical
depression levels on the SOl or other measure at any time past the
intervention.

199



The issue arises of a possible need for follow-up training by which an
individual can revise or sharpen CST skills and their application. This
would be particularly relevant where individuals have not experienced
sufficiently negative events to have learnt the application of CST skills in
enough different situations in order to generalise learning. That is, in line
with the view by Seligman (2006) they aim to maintain their wellness
rather than wait until they are facing the possibility of a relapse.

(c) Third research question
The third research question, whether the intervention is more likely to
retain clients through to completion compared to other interventions does
not seem to find support from the results of this study. This finding is
consistent with the literature. A comparison by the Fisher Exact test
between the ratios of completers to participants of face-to-face treatment
studies by Cahill et al. (2003), Westen & Morrison (2001), Shapiro et al.
(1994), Jacobson et al (1996) and Jacobson et al. (1991) finds no
significant differences (p> .05).

The essential difference between the comparison studies and the present
study are that in the present stUdy participants and provider maintain a
close therapeutic relationship throughout the intervention by email and
face-to-face where needed by the participant, and this personalisation
tends to encourage continuation on the intervention. In the cases of the
three individuals who did not complete their treatment the reasons were
simply that their depression had receded. In two of these cases the parts
of the course they completed provided them with sufficient impetus to
take charge of their own treatment without further help.

(d) Fourth research question
The answer to the fourth research question, namely the size of the
potential savings of this intervention compared to other interventions, is
62%. Assuming that further studies with appropriate control groups
validate this finding a reduction of this magnitude presents a significant
saving in treatment costs and thus maximises the value of the Medicare

200



benefit for psychological treatment. At the same time it frees up
psychologist resources to be used to increase throughput. The savings
generated from psychological interventions were estimated at 200/0 in
meta analysis of 91 studies by Chiles, Lambert and Hatch (1999) which
were published between 1967 and 1997 using meta-analytic techniques.
The intervention lends support to the view by Ahearn et al. (2006) that
eHealth is emerging as a vehicle to shore up the inadequacy of the
healthcare system to meet needs for chronic disease management.

(iv) Limitations of the study
The study has several limitations including:

• The availability of the intervention is limited to depressed
individuals who, apart from not having psychotic symptoms, have
(1) access to the Internet, (2) the requisite information technology
skills, (3) adequate English literacy to understand the information
presented and (4) the ability to compose and communicate with
emails. Those without computers and with low English reading
and writing skills are less likely to be able to participate;

• the small size of the sample. This reduces the validity of
comparing results with other studies and also generalising the
findings. The best that can be said is that the study points to the
need to undertake larger scale studies with control groups;

• the lack of randomization. The sample was purposive. It was not
chosen on a random basis. Hence its capacity to be generalised
cannot be assessed mathematically;

• there was no control group. Hence there is no evidence that
improvements noted could not have occurred because of the
passage of time;

• there was a "power" relationship between the writer delivering the
program and the clients receiving the intervention. However the
implication which was understood by all participants was that if they
did not deem their progress in addressing depression to be

201



satisfactory they could leave the program and seek alternative
treatment through their GP

• the choice of a three-month follow-up period. Although this can be
argued as sufficient, a longer fOllow-up period would have likely
been able to test the ability of the individual to apply the skills
learnt in the intervention in a wider range of life situations than
those likely to be encountered within three months;

• the restriction of participants to those with mild and moderate
depression. There is insufficient evidence as yet to support the
efficacy of CST for severe depression;

• the lack of evaluation through Likert Scales or comments of the
specific treatment outcomes of adding supports to the online
course such as online sessions or emails.This would have been
too difficult to obtain because of the integrated nature of the
course. Many sessions were used for remedial work and hence
their availability would have been crucial to those who needed to
have them. Emails were very much a part of the course and
participants would not have validly been able to estimate the effect
of each to their treatment.

(v) Recommendations for further research
The present study researches the concept of treating depression with an
intervention which integrates a course in CST skills with face-to-face
sessions. The sessions aim to help the individual clarify any difficulties in
learning CST concepts and apply the skills learnt in the course to address
the disorder. The concept requires more research with statistically
random and larger samples utilising control groups before it can be
accepted as an effective approach in the treatment of depression.

A suitable study would be allocating at random volunteers for treatment
to wholly face-to-face sessions and to an integrated program of sessions
and course designed and delivered by the same provider. There is

202



considerable preparatory work required before such a study can be
validly undertaken. For example treatment providers need to learn
website development skills and be supported whilst developing their sites,
hosting capacity has to be acquired and sites uploaded and tested.

Similar studies are needed which include clients at the severe level of
depression. A study by Prisciandaro and Roberts (2005) showed that
depression, as defined by the symptom-focused DSMIV, was best
considered as a dimensional construct. That is the categories of mild,
moderate and severe in the BOI or other tests are artificial dividers on a
continuum rather than denote real differences between the depression
categories. If this is the case there is no valid reason for not expanding
studies to include individuals with depression across the range of the
continuum.

Treatment modalities need to be established by which follow-up or
refresher training in CBT skills can be effectively delivered. Modalities
might include an online intervention either the one completed or a
different one such as MoodGYM, attending one or a small number of
sessions according to need or both. Bibliotherapy could also playa role
because it would provide an individual with a different approach to CST
thereby enriching learning.

The practicability should be researched of refining online intervention
measures to cater for individual differences in learning rates and time
availability for the intervention. Individuals might show sufficient
improvement to address their disorder within a small number of sessions
and so lose their motivation to complete the course, as has happened in
the present study. On the other hand safeguards are needed to ensure
that temporary 'placebo'-like improvements are not mistaken for robust
change and lead to the participant deciding to end the intervention. This
issue is less likely to arise where face-to-face sessions are interspersed
with the online course because the treatment provider can alert the
individual to the likelihood that this might be occurring.

203



The selection of participants who are best suited to treatment with an
online component needs to be based on criteria established by research.
There is a self-selection element involved in that individuals without
knowledge or a liking for online work are less likely to opt for this
modality. However there is no certainty that just because an individual
feels at ease in an online environment he or she would do better in an
online intervention. An approach is needed to provide individuals with
guidance to make effective choices in treatment modality and providers to
assist them in making these choices.

The concept of stepped care was often cited in this study. Research is
needed to establish the most effective treatment modality for given levels
of severity of depression. The writer envisages that a client would
complete the BDI, undergo a clinical assessment to validate the test
score and be recommended a form of treatment on the basis of score,
clinical judgement and client wishes. For example clients scoring at
subclinical levels would be recommended inspirational readings and
video presentations. At progressively higher BDI scores and validating
assessments clients would be invited to consider bibliotherapy, online
treatment in a self-help site, online and face-to-face as in the present
stUdy and for the most severe forms intensive face-to-face therapy. This
approach does not deny the individual's right to choose medication or
other forms of treatment. Some of the writer's clients did opt for
medication with a psychiatrist when they perceived they were not
progressing sufficiently well. Most returned to CBT when they realised
that medication and side effects were interfering with their functioning and
also, for some, because its interaction with alcohol removed an important
mode of relaxation they shared with their friends and families at
mealtimes. Where appropriate the writer enjoys advising his clients that
CBT and a glass of wine are compatible.

The present study is on depression. Although this disorder deserves
most attention given its high prevalence in the community other disorders

204



including panic, generalised anxiety, the phobias, bulimia and anorexia
also deserve attention. CST is a treatment methodology that is highly
amenable to self-management as demonstrated by Internet, bibliotherapy
and manualised face-to-face studies. Similar control-group studies
should test the efficacy of integrated course and session CST
interventions in different disorders. Providers intending to utilise this type
of intervention could develop the core of CST in one module with other
modules covering the specifics of individual disorders. In this way
providers could maximise the numbers of clients treated. One of the
advantages of the integrated intervention is hypothesised to be that
participants become valuable contributors to continuous development of
the online course and the approach itself. The provider is able to collect
data from every client and use the information to tune the course and the
intervention closer to needs.

The writer is aware that non English language communities have a dearth
of psychologists who are bilingual to the extent that they can deliver
treatment in a language other than English. The study points to a
possible venue by which older people in these communities can access
treatment. Although they might not have adequate online skills they likely
have children or grandchildren with the skills who would be willing to
provide some elementary training and access time to their own
computers. Following completion of this study the writer intends to
prepare a similar site in Greek because through his practice he has seen
many older persons from this background with Internet access at home
and a willingness by children and grandchildren to help their old folk
explore this medium.

Motivation to complete treatment is an important factor for further
research. The less engaging the treatment the more likely that the
individual might abandon it. The Internet course appears to have in-built
attractions for clients. It is self-contained and from the start clients know
the amount of reading material and exercises it includes. They are
therefore in a position to plan, which itself is a useful skill in overcoming

205



depression. This leads to an enhanced sense of responsibility and
ownership - they are in charge of their treatment. They are not entirely
dependent on someone else to show them the way out of their
depression. Another factor is that some clients with depression might
miss sessions due to low mood and other impediments. Because the
course is online participants are able to pursue learning when they feel
best able to do so. Postponing sessions whilst on the online course is not
necessarily fatal to progress and any queries can be managed by email.
The requirement to undertake a task, especially of a cognitive nature is
often perceived as daunting by a depressed individual. However this is
offset by having to attend fewer sessions in the beginning, before
improvement begins, when motivation is low.

One of the assumptions of online treatment that merits research in
relation to its contribution to the treatment process is the inherent ready
availability of enrichment material. No live treatment can provide a client
with the richness of material on the Internet. Participants can find
different explanations for concepts and examples to supplement their
learning or to help them appreciate a topic in its entirety. The therapist
acts more as a coordinator and facilitator of learning rather than as a
dispenser of knowledge. Training in skills is not as time constrained as in
face-to-face treatment because it is undertaken by the participant working
on the online course at his or her optimum pace. The sessions provide
the opportunity to expand CST skills training into coaching and shaping
behaviour in a range of life situations confronting the client. The proviso
that individuals have online skills does not appear to be a daunting
obstacle. In the study, one completer had rudimentary skills in accessing
the Internet. He was briefly shown how to access the site and progress
with online treatment. He became adept in a short time. All these
factors should be researched to determine the validity of their contribution
to the treatment process.

The savings demonstrated by the study is a key advantage in today's
strained government health budgets. The study shows that for most

206



clients the Medicare rebate limit of twelve sessions annually is likely to be
sufficient for treating all but those who have a lengthy history of
depression. For the latter there is always the possibility of continuing
their treatment online through sites such as MoodGYM. Where money is
not a pressing concern or the individual has private health insurance to
supplement Medicare, continuing to attend face-to-face sessions with the
same or different provider is an option. In any case, individuals should
have continuing access to the online course or a publicly available
intervention such as MoodGYM and the option of pursuing material in
referred sites.

The issue of non compieters appears to be less relevant to the success of
an online treatment program than anticipated at the beginning of the
study. Online treatment has the inherent capacity to cater for individuals
with widely differing learning abilities, preferences and time availability. In
the study the non completers did not abandon the course because of
difficulty or loss of interest. One was motivated to pursue a different
approach to overcoming his depression, another simply read ahead and
applied the skills without treatment provider support and the third
experienced a change in life circumstances. Interventions using online
courses should be designed flexibly to cater for the range of individual
differences and particularly to avoid implying that depression won't lift
unless all segments of the course are completed.

Charging clients for the online component in an intervention which
combines both face-to-face and online is a challenge. The writer has
made the site freely available to his clients and has not charged them.
However it is possible that a client attending the introductory session
during which access to the site is arranged might find as he or she
progresses with online treatment that this is enough by itself to lift
depression. Unless some form of charging is introduced the provider is
not being recompensed. This issue is made more difficult to resolve
where Medicare or the private health funds are involved, which pay only
for treatment directly provided to the client.

207



(vi) Final comment
The expansion of healthcare provision through the Internet is unlikely to
wither. There is a need to research improvements in intervention delivery
and at the same time to overcome the many impediments before findings
in studies such as the present one are able to make an impact - the
profession needs to be convinced and professionals require training and
support in website design and online treatment delivery. Ethical
considerations should be addressed and guidelines written.

Online interventions are likely to attract individuals who are
knowledgeable about the Internet and are comfortable with taking charge
of their own treatment. Given the current and the projected levels of
access to the Internet, interventions such as in the present study are
becoming available to an increasing proportion of the population. There
are some imbalances such as the lower levels of access by older people
who tend to experience depression and other psychological ills more
frequently and by the economically deprived. The writer believes that the
proportion of older Internet users might be increasing at a greater rate
than estimated. He is president of a group of former RAAF national
servicemen in Sydney, all close to the age of 70. Of the 60 members
who have joined up to November 07,35 or over half have Internet access
and most have broadband. About half have acquired online access in
the last two years.

The study shows that the therapeutic benefits available through online
treatment are not confined to sites especially designed and set up to
meet needs of the Internet community. A practitioner with training and
experience in face-to-face counselling and some computer knowledge is
able to develop a website to take over some of the information giving and
skills development in the treatment of depression by CBT. There are
software packages that made website design an easy process, such as
Microsoft FrontPage, Microsoft Publisher, Broderbund PrintShop and A4

208



Desk which uses Flash for designing animated presentations. PrintShop
is low-cost software while A4 Desk can be freely downloaded.

Online work by the clients can be fully integrated into the treatment plan.
Apart from explaining any concepts which the client finds difficulty in
understanding whilst undergoing the online course, sessions are devoted
to ensuring clients are able to apply the skills learnt to their day-to-day
functioning. Being online the course provides ready access to all topics
that need coverage and hence is sensitive to individual differences in
learning rates, time availability and motivation to pursue enrichment by
accessing other sites with related topics. The perceived and actual
availability of the live therapist creates a sense of partnership. Sessions
help to establish a healthy and vibrant client-therapist relationship. The
routine aspects of teaching information and skills are pursued by the
clients themselves through the course. Skills coaching, that is applying
CBT to the resolution of real life issues confronting clients, is done in the
sessions. Apart from leading to a reduction in face-to-face sessions this
approach cements the client-therapist relationship into one of a
partnership. The writer normally requires about eight to twelve sessions
to treat depression for moderately depressed clients. The stUdy took an
average of less than five. There were however large variations, from ten
for a male in his late 40s with a history of lifetime depression episodes to
two for two males who only needed two face-to-face sessions. The writer
estimates that he would have needed a minimum of 16 hours to treat the
first participant hence the Internet course did result in cost and time
savings.

The study has shown that the argument that is putting online and face-to­
face treatments at opposite ends is nonsense. Both have their strengths
and roles to perform. An approach that integrates both modalities into
one intervention draws benefits from each, reduces costs and treatment
times and importantly puts clients in charge of their own treatment.

209